bookworm/bookworm-smart-assistant
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Compare commits

base: bookworm:v6.7.0
Branches Tags
bookworm:main
bookworm:v6.7.0
bookworm:v6.5.1
...
compare: bookworm:main
Branches Tags
bookworm:main
bookworm:v6.7.0
bookworm:v6.5.1

2 Commits
v6.7.0 ... main

Author SHA1 Message Date
Bookworm Admin
34f304881f fix: strip session-continuity-mcp hooks from Portable template 
export.mjs now removes hooks referencing npm packages not included
in the Portable distribution (session-continuity-mcp).
Eliminates MODULE_NOT_FOUND errors on Portable installations.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-27 22:15:39 +08:00
Bookworm Admin
131840c962 fix: route-interceptor fail-open + v6.6.0 sync 
- route-interceptor-bundle.js: 硬 require 改为 try-catch fail-open
  (旧版部署缺 scripts/route-engine.js 等文件时不再崩溃, 降级为 BWR:skip)
- package.json 6.5.0 → 6.6.0 对齐 VERSION 文件
- INTEGRITY + 签名更新

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-27 19:30:24 +08:00
136 changed files with 79049 additions and 113104 deletions
Show Stats Expand all files Collapse all files

4
CLAUDE.md
View File

@ -1,4 +1,4 @@
# Bookworm Smart Assistant - 智能路由系统 v6.6.0-phase1-B # Bookworm Smart Assistant - 智能路由系统 v6.6.1
## 会话激活横幅 ## 会话激活横幅
@ -35,7 +35,7 @@
5. **候选回退**: 主路由不适合时可从候选列表选择 5. **候选回退**: 主路由不适合时可从候选列表选择
6. **默认回退**: developer-expert 6. **默认回退**: developer-expert
> 消歧规则由 hooks 自动应用，完整 89 条见 scripts/disambiguation-rules.json > 消歧规则由 hooks 自动应用，完整 93 条见 scripts/disambiguation-rules.json
--- ---

154
INTEGRITY.sha256
View File

@ -1,22 +1,22 @@
36eb7e81651773c508b5d9c28d3a70b40cc8a77594363331c0530710c3098d88 agents/canvas-ui-designer.md 36eb7e81651773c508b5d9c28d3a70b40cc8a77594363331c0530710c3098d88 agents/canvas-ui-designer.md
6d1ae5ee44805406ebb22380385fc156899b9a6b3f28c80ccc407eda65a3f6a1 agents/code-reviewer.md 507151dc35508692053a479f4997214cb456a23f89e42820da74163992935c00 agents/code-reviewer.md
472d5a49449a9640871e081ef1ee3943b2f9b9edc5c0dacd7221329a6be451f3 agents/delivery-quality-assessor.md e1569cf94896a62aec8ffe9478cff84f109feeb5dc22d38af9575f7902ec8f72 agents/delivery-quality-assessor.md
0cf2a455a7064b2ee8ff39ca2ff81a84f323f3ca0a67fbb1cf41f12368857bde agents/desktop-automator.md 0cf2a455a7064b2ee8ff39ca2ff81a84f323f3ca0a67fbb1cf41f12368857bde agents/desktop-automator.md
c341004ee55af06d854815d42ed6a90e8a9d18859a70264c17b52d0a7f3f8271 agents/explore.md c341004ee55af06d854815d42ed6a90e8a9d18859a70264c17b52d0a7f3f8271 agents/explore.md
4dd91dd220b4800747b06dd3bf07c600d62865e2a200925189c44a2581e7010d agents/full-stack-builder.md 4dd91dd220b4800747b06dd3bf07c600d62865e2a200925189c44a2581e7010d agents/full-stack-builder.md
6a58612c190a60fc7602b6fc3d2a233878e252a9cdd389839c8676276ff2df08 agents/module-integrator.md 6a58612c190a60fc7602b6fc3d2a233878e252a9cdd389839c8676276ff2df08 agents/module-integrator.md
90b7dd397f8f83d2158b3671871faa4cb08ce7511e2a568a3a819e3bfe4803ba agents/orchestrator.md 6905ff9a04228e6aceeb3e07b6cff39eb283ae8c00cd5e675ad1bf3494e59a99 agents/orchestrator.md
54fa0a82ad21045ea331b127d35bf6fc14ee29c5cbab78689e15a7381da02460 agents/pre-deploy-checker.md 54fa0a82ad21045ea331b127d35bf6fc14ee29c5cbab78689e15a7381da02460 agents/pre-deploy-checker.md
86b5e4ec27f9c5d020071b5cd98996ef0e4eba7928cf3f2d225033687d7d9ce8 agents/production-reviewer.md 86b5e4ec27f9c5d020071b5cd98996ef0e4eba7928cf3f2d225033687d7d9ce8 agents/production-reviewer.md
6da2f9a9e34b07bbdb7494b9748da2e91bea6ac3d74a372a599683d9b8bc73a1 agents/quality-gate.md 6da2f9a9e34b07bbdb7494b9748da2e91bea6ac3d74a372a599683d9b8bc73a1 agents/quality-gate.md
06b4f6882cf87e512653533f36b1c356580cfe73fc494bd3b12690229a507d62 agents/red-team-attacker.md 06b4f6882cf87e512653533f36b1c356580cfe73fc494bd3b12690229a507d62 agents/red-team-attacker.md
341e660a37ca6330e017ca48d2147c16eb9f751a0066f0f3a8c3f3eeadbf0777 agents/red-team-logic.md 341e660a37ca6330e017ca48d2147c16eb9f751a0066f0f3a8c3f3eeadbf0777 agents/red-team-logic.md
b3b64d847cbb8e081de113097d79ecc4101f56b226b9a52d7bcc1117a389b5df agents/research-analyst.md b3b64d847cbb8e081de113097d79ecc4101f56b226b9a52d7bcc1117a389b5df agents/research-analyst.md
f74c84610bfc163b4f985bb8ad34f68096e852b5639b0e1dad0ec0caa317cf32 agents/security-hardener.md 137b584f734f5fdf487f6518193a9ef96712ad70a73831f6d3e7d984f587f51d agents/security-hardener.md
84043af52f098b7e4daf48c265f23932053e1f442c96daeb675bdd65a8283966 agents/self-auditor.md 84043af52f098b7e4daf48c265f23932053e1f442c96daeb675bdd65a8283966 agents/self-auditor.md
70c209872f94be2eb48e2659fbeb93dc89007002848990ebb8654f469ed70bc9 agents/self-healer.md 70c209872f94be2eb48e2659fbeb93dc89007002848990ebb8654f469ed70bc9 agents/self-healer.md
f3c4467485e1b7f785aaf802fcd75a663601389434d6b6aef87b476fa4e63aea agents/test-writer.md f3c4467485e1b7f785aaf802fcd75a663601389434d6b6aef87b476fa4e63aea agents/test-writer.md
de3d4906c6d4fe902efeaec4f76ecefcc8ec17f0abe1243c318bc90608dcf333 CLAUDE.md e12ddb007ff40a7449500bfb9f801d68eb137268401afc4872c751187ce2ddf5 CLAUDE.md
5774b2396bd2e032d1414d5030e047361676906b4384d9e3956d3ec3ced42924 constitution/AI-CONSTITUTION-CORE.md 5774b2396bd2e032d1414d5030e047361676906b4384d9e3956d3ec3ced42924 constitution/AI-CONSTITUTION-CORE.md
d3c228e22e05a1ca88c38cab5d0cf1f36104646bbf74a44d1c683e760a864351 constitution/AI-CONSTITUTION-PRODUCT.md d3c228e22e05a1ca88c38cab5d0cf1f36104646bbf74a44d1c683e760a864351 constitution/AI-CONSTITUTION-PRODUCT.md
6b9de5a39fbc3afbd0c44f0488785a585d3ba6a192e7e995a2a4545fdb1fc9c3 constitution/AI-CONSTITUTION.md 6b9de5a39fbc3afbd0c44f0488785a585d3ba6a192e7e995a2a4545fdb1fc9c3 constitution/AI-CONSTITUTION.md
@ -74,8 +74,8 @@ d06c74f7e21ef294f1fd1a1f2d5d8eb4f4b3d9e2769550c8532970d99033c1cc hooks/block-se
9c18207c864eb42224d208ff2871de61c0bfe5843efd35f6e0a766c8f0079942 hooks/check-gray-expiry.js 9c18207c864eb42224d208ff2871de61c0bfe5843efd35f6e0a766c8f0079942 hooks/check-gray-expiry.js
6d2b39448407b05ada21af262d3805580fb312ed3146e16e3c070c40e618f14d hooks/check-lint.js 6d2b39448407b05ada21af262d3805580fb312ed3146e16e3c070c40e618f14d hooks/check-lint.js
533572b00c290b21f970608d8778e7201dcaff126a38a956529f716fc2b6b265 hooks/check-typescript.js 533572b00c290b21f970608d8778e7201dcaff126a38a956529f716fc2b6b265 hooks/check-typescript.js
a382a8c7fcadaf1b57eaecdcdfd7a6de8554b1d6f14687cca5f03262153522b5 hooks/checksums.json 36db194abe857b0585e10f9f691d3add4be6f9d346ece58cd084884c6f20aba0 hooks/checksums.json
725e75e353a9d866bc2c3609cb85bcf404ccd99059197a7d56167612aec0081f hooks/checksums.sig 286d9cf47136ac58df4be19d7ac7759d5e49d3180f801c877ffa77a422aaedee hooks/checksums.sig
1b1e7fab96e25760b94eaa935529920f1ab7d38b2b231ee4642bae99465109b2 hooks/clipboard-image-hook.js 1b1e7fab96e25760b94eaa935529920f1ab7d38b2b231ee4642bae99465109b2 hooks/clipboard-image-hook.js
dec6ceb0da432bef7de941fe92ea412ba116aa3143765b904fcdd4d691a0ff83 hooks/code-quality-gate.js dec6ceb0da432bef7de941fe92ea412ba116aa3143765b904fcdd4d691a0ff83 hooks/code-quality-gate.js
2c7441e6ea9a2704f7534156a0c1f7879405ee0bccf976690585480563caa04c hooks/commit-message-lint.js 2c7441e6ea9a2704f7534156a0c1f7879405ee0bccf976690585480563caa04c hooks/commit-message-lint.js
@ -88,7 +88,7 @@ c8c503b2d6b94464f9f9f9f91517a8cc33d0bbc8f43bc5cce73589631d3c91a5 hooks/context-
2690c97401b6f913c0e90d6f49349e133cdc7d81faa61f6cd865f00e9ed58965 hooks/edit-precheck-dispatcher.js 2690c97401b6f913c0e90d6f49349e133cdc7d81faa61f6cd865f00e9ed58965 hooks/edit-precheck-dispatcher.js
a53a623e49c8384483ef076f3513c42de55c46217ef6506c4c1fb2c692b0f3f8 hooks/integrity-check.js a53a623e49c8384483ef076f3513c42de55c46217ef6506c4c1fb2c692b0f3f8 hooks/integrity-check.js
121b3de35541f7bff7624e411c9b496282444f99d64cc5dff73c15a0c9afd9ea hooks/integrity-check.js.self-hash 121b3de35541f7bff7624e411c9b496282444f99d64cc5dff73c15a0c9afd9ea hooks/integrity-check.js.self-hash
6ff44a5e8427fde8024152ddb5680093875731fb8da0757ead6c7ba1de989570 hooks/lib/fail-mode.js 8d7e4508dda23416f454e4e4571b8e0cc76ebae467c43b785647ff072d83c9c9 hooks/lib/fail-mode.js
58d6ef4ffa50f69944d43b5551cc6f8adfe84de3166defac1d443a6083957bfc hooks/lib/fast-cache.js 58d6ef4ffa50f69944d43b5551cc6f8adfe84de3166defac1d443a6083957bfc hooks/lib/fast-cache.js
0683f0c43f65c80a159ef700f7f975b07f2187df61f08130099bf582a6486c44 hooks/lib/jsonl-hmac.js 0683f0c43f65c80a159ef700f7f975b07f2187df61f08130099bf582a6486c44 hooks/lib/jsonl-hmac.js
63792b207b6148ddc6bb2a8a2a24e3309cfc3e66b4773d4eeed9d811da3ef997 hooks/lib/metrics.js 63792b207b6148ddc6bb2a8a2a24e3309cfc3e66b4773d4eeed9d811da3ef997 hooks/lib/metrics.js
@ -121,7 +121,7 @@ f5d33a50c89f30c0596f152452989dd5f3214829aed2b2bfbd7176537175e424 hooks/project-
fe01f02e0d45b4d5f8890c226a6433d0dcc6f68932463ecbed3df1ba12ac0a09 hooks/rollback-on-fail.js fe01f02e0d45b4d5f8890c226a6433d0dcc6f68932463ecbed3df1ba12ac0a09 hooks/rollback-on-fail.js
f584ab2984108bcf2ff5179bbb4472e3c4b8b5bb86fcd0c15cb0b82649732323 hooks/route-auditor.js f584ab2984108bcf2ff5179bbb4472e3c4b8b5bb86fcd0c15cb0b82649732323 hooks/route-auditor.js
8caf07ee3ee4df370e058dbf54a3a91450eb5ba63fe672051ede2e2d29dc30c0 hooks/route-compliance-gate.js 8caf07ee3ee4df370e058dbf54a3a91450eb5ba63fe672051ede2e2d29dc30c0 hooks/route-compliance-gate.js
792cb71f4e4379aac7faaf1c00e05f4eb65e1db7052ab7b38d2c47ec02e81db1 hooks/route-interceptor-bundle.js b9a88af4576f3dcb04e49fcf849721e8163dae052197b76e9419fece52c38e32 hooks/route-interceptor-bundle.js
cdaacb861d6c889d38e82fbd172c369d4303fc6a904240ac882d0cbae4ca691f hooks/route-interceptor-bundle.js.bak-p21.1777282215104 cdaacb861d6c889d38e82fbd172c369d4303fc6a904240ac882d0cbae4ca691f hooks/route-interceptor-bundle.js.bak-p21.1777282215104
16dd875fee2994e26dd570bb49b4399b075fd02e5e23e67d3ec775fbf62a83a3 hooks/rules/ask-patterns.json 16dd875fee2994e26dd570bb49b4399b075fd02e5e23e67d3ec775fbf62a83a3 hooks/rules/ask-patterns.json
98f490fb78c7567b8709570aaff66f1ce10c45522e4fa3b2a3cf8b4a0991ec17 hooks/rules/credential-patterns.json 98f490fb78c7567b8709570aaff66f1ce10c45522e4fa3b2a3cf8b4a0991ec17 hooks/rules/credential-patterns.json
@ -155,7 +155,7 @@ d6b0024d7c0bd2684cb341584697548cfa5ae555ccdeb0be8b4c8f100407f28d hooks/token-sa
ede508bf85863988c6f205a89315c69303373b81879832302af3a1b10634e634 lib/activate.js ede508bf85863988c6f205a89315c69303373b81879832302af3a1b10634e634 lib/activate.js
cba84f84fa95bc61cb7137734337efd784de26528b503c822aff654073d6267f lib/fingerprint.js cba84f84fa95bc61cb7137734337efd784de26528b503c822aff654073d6267f lib/fingerprint.js
692c6ebb5c199961e697156c4cc2c0ba0b183ff0436573a902d6b6eeea29118f lib/load-skill.js 692c6ebb5c199961e697156c4cc2c0ba0b183ff0436573a902d6b6eeea29118f lib/load-skill.js
867af94a38d207384e36ffd6b76e3758b8e76f003a100cac588b856ff78f3de6 package.json e407ee5776a08241ccd98812dfcec0b2b9d94b944f3fba7f82c1e4aa03a44394 package.json
d5cc0c072ddab9ea1f5cc8a13361531170d01e9043c9e251042833b2b15be91f scripts/ab-backtest.js d5cc0c072ddab9ea1f5cc8a13361531170d01e9043c9e251042833b2b15be91f scripts/ab-backtest.js
8271d2e8c369dc039af32e713bc96464caeef31f28758543f68d8dd49a5729c0 scripts/adaptive-disambiguator.js 8271d2e8c369dc039af32e713bc96464caeef31f28758543f68d8dd49a5729c0 scripts/adaptive-disambiguator.js
b99aa8993a46a4f57c32a19650244a2ae100f64203f914d94c547e946851e10b scripts/add_css_patch.py b99aa8993a46a4f57c32a19650244a2ae100f64203f914d94c547e946851e10b scripts/add_css_patch.py
@ -211,7 +211,7 @@ c3a0f0ad8050e9b875fc0aee4deac11f16dd83af7f4350e6934fa0f8d8ea5e3d scripts/dashbo
4e5492032e27e3c0d5c3d5e7c943daa175883ebd7ed43d67185a3241420596aa scripts/deploy-portable.js 4e5492032e27e3c0d5c3d5e7c943daa175883ebd7ed43d67185a3241420596aa scripts/deploy-portable.js
be654fc0a5adbe51ac546bbf164bcc6db6ded5c536389bbe2be2befbc8e7fa1e scripts/deploy-transactional.sh be654fc0a5adbe51ac546bbf164bcc6db6ded5c536389bbe2be2befbc8e7fa1e scripts/deploy-transactional.sh
77abe264191760b2d46e919f4a4fd4e345cb573bbdd8ccd88e654aa74b5dc894 scripts/deterministic-quality-gate.js 77abe264191760b2d46e919f4a4fd4e345cb573bbdd8ccd88e654aa74b5dc894 scripts/deterministic-quality-gate.js
8b75e8f538af92f61371d94672d07ff874441ef91badfc3939e53a515e2892c3 scripts/disambiguation-rules.json 18316935296758d4e5d3bcdd466c29019d221b0c37260f2c790ae7ef8e13b9e6 scripts/disambiguation-rules.json
56ed2c18a52d613e3f77156d049337d7c0cd24f3dcb8eb1e54977983f5e4ba56 scripts/disambiguation-rules.json.backup-t5-2026-04-16 56ed2c18a52d613e3f77156d049337d7c0cd24f3dcb8eb1e54977983f5e4ba56 scripts/disambiguation-rules.json.backup-t5-2026-04-16
a744e559196f1da6528760e94d4b85b398ab5760ba98b9421a2804ba36bd7fec scripts/disambiguation-tree.js a744e559196f1da6528760e94d4b85b398ab5760ba98b9421a2804ba36bd7fec scripts/disambiguation-tree.js
269556f0e1baf2a7a72cae6ce0d79e42b30d42f2d56c71de0f641920b3a0d339 scripts/domain-capacity-manager.js 269556f0e1baf2a7a72cae6ce0d79e42b30d42f2d56c71de0f641920b3a0d339 scripts/domain-capacity-manager.js
@ -222,7 +222,7 @@ ea0c5a066a5a79137acdc075475860c1182f53cafb1c94c9db826284c0e55145 scripts/embedd
088dc672368ef9f59c64a01613653a6ae755f9c698d3a731fa9cc5cda75f79b1 scripts/fusion-weight-learner.js 088dc672368ef9f59c64a01613653a6ae755f9c698d3a731fa9cc5cda75f79b1 scripts/fusion-weight-learner.js
87f26a1c0ca112dbd9eb4e2cf0126b12d7832c5c2320c23c48406c07bdb974af scripts/gen_git_ui.py 87f26a1c0ca112dbd9eb4e2cf0126b12d7832c5c2320c23c48406c07bdb974af scripts/gen_git_ui.py
ada60a5af89c652a3cfb7dd2986189a363814bd236bfc645ab600494a178932a scripts/generate-skill-index.js ada60a5af89c652a3cfb7dd2986189a363814bd236bfc645ab600494a178932a scripts/generate-skill-index.js
724f3a221c67b31d044268325c53dd799fdcf95db13d4b9e48d2252389ca81b6 scripts/generate-stats.js 7dfec0c1f3b09386ac7a41d1804dee80e108a1d613dc99ccf73935984fa7c7a6 scripts/generate-stats.js
2d3f2f1f0c6a3556fafd9beacd0ebc69367fed9816dd33fba1a0daaad86af0fd scripts/golden-set.json 2d3f2f1f0c6a3556fafd9beacd0ebc69367fed9816dd33fba1a0daaad86af0fd scripts/golden-set.json
ab1e7b60def2f01ea83419af1caa9d24e8cc62f142bb377aa9199d5f15e97b88 scripts/health-check.js ab1e7b60def2f01ea83419af1caa9d24e8cc62f142bb377aa9199d5f15e97b88 scripts/health-check.js
7be805736c54917c19a8249c360a3e17e9c023ad72956a06d67f6f5fb4120a69 scripts/hook-priority-scheduler.js 7be805736c54917c19a8249c360a3e17e9c023ad72956a06d67f6f5fb4120a69 scripts/hook-priority-scheduler.js
@ -241,8 +241,14 @@ a3d16d0c5be8ca18f6eee20a02258225bb7630fa2b220cedef867af0025e6998 scripts/mcp-us
9802e3f56c786e4cfddf93e69ea6247759b85b1ac92cd6194ffaf6a35f872bdf scripts/multi_ai.py 9802e3f56c786e4cfddf93e69ea6247759b85b1ac92cd6194ffaf6a35f872bdf scripts/multi_ai.py
f1a5c0a5df7f656e4c2c03788fde38f2d51551dd60b4102b2cb9e008cb6593f9 scripts/patches/_observer-summary.js f1a5c0a5df7f656e4c2c03788fde38f2d51551dd60b4102b2cb9e008cb6593f9 scripts/patches/_observer-summary.js
b748c0bb80bc513ca68468208ce9ddec639b5e3b34f50059242e8b5ed5659de2 scripts/patches/_observer-tests.js b748c0bb80bc513ca68468208ce9ddec639b5e3b34f50059242e8b5ed5659de2 scripts/patches/_observer-tests.js
7faefd8859e4bc8623dcd052b173bff39b6a897fd59423253e67c08052385488 scripts/patches/bump-v6.6.1.js
e0fc9f92eba1fc00e71b82899b5129c5fd294c51fb2942e9d1d4ad4dd2137efc scripts/patches/debug-evolution-log-line55.js e0fc9f92eba1fc00e71b82899b5129c5fd294c51fb2942e9d1d4ad4dd2137efc scripts/patches/debug-evolution-log-line55.js
07d251955cb00d3cf8a76080d9193ce54ea862d8d68cf40704acdd05d9552516 scripts/patches/fix-cold-cap-override-0427.js
0402c25c16fef1d4c2a596ab8da76019a12226a7f45db6c0bef5fed2ef72a9c7 scripts/patches/fix-composable-regex-0427.js
b062554f6f07cb3fbcc9356d5cc5148fcb62c1cf4e9c1b9bf4b8e3e5ff98aebe scripts/patches/fix-lvp-persist-0427.js
e6896b06a915e5cc7eae9829efdfa2b32b96d1d31e863c485952b227552f0600 scripts/patches/fix-w1-w5-audit-0427.js
b7d299d60785af25f75710a622eb84d1dda3cb988fef060eec802c0f3f400be0 scripts/patches/install-task-scheduler-verify.cmd b7d299d60785af25f75710a622eb84d1dda3cb988fef060eec802c0f3f400be0 scripts/patches/install-task-scheduler-verify.cmd
1a7d05050bc5c20b12ca9c185ef7b4e52be3ab7fe1fc25582f7d45a3529497c6 scripts/patches/migrate-session-continuity-to-local.js
13a844fdefce12de6fae01b51ccd92cbc6a1c9531470bc42bdd0b30b44936c6e scripts/patches/patch-add-staging-pipeline-flag.js 13a844fdefce12de6fae01b51ccd92cbc6a1c9531470bc42bdd0b30b44936c6e scripts/patches/patch-add-staging-pipeline-flag.js
dcc42834bbec89a692b69cd3b06aa1c8f3dc868b20fcf5ca843432839b8a25c4 scripts/patches/patch-audit-fix-registry-drift.js dcc42834bbec89a692b69cd3b06aa1c8f3dc868b20fcf5ca843432839b8a25c4 scripts/patches/patch-audit-fix-registry-drift.js
9ccb6cb45c56fe3436a5786446bb89d48ef5d94cdf061d024daae9a22d596e7a scripts/patches/patch-banner-route-accuracy.js 9ccb6cb45c56fe3436a5786446bb89d48ef5d94cdf061d024daae9a22d596e7a scripts/patches/patch-banner-route-accuracy.js
@ -270,6 +276,7 @@ fd3556fc00c973e69759bdacbf4c4ac117793414c729b29c8c80e3baad8e6f9a scripts/patche
8f0a5ffd480047b47f156350a856a3a013faec0956f60302c57edc95c6ba1abc scripts/patches/patch-memory-audit-snapshot.js 8f0a5ffd480047b47f156350a856a3a013faec0956f60302c57edc95c6ba1abc scripts/patches/patch-memory-audit-snapshot.js
fd60eb6872350e6dbf505364ba46deaa49b39b90879a6b7d20270c5314f707d0 scripts/patches/patch-p0-1-metrics-emit.js fd60eb6872350e6dbf505364ba46deaa49b39b90879a6b7d20270c5314f707d0 scripts/patches/patch-p0-1-metrics-emit.js
42fa8628b9c5890bd371063dc2708cfe18a97675b98493aeb439c29bb941db38 scripts/patches/patch-p0-2-session-once.js 42fa8628b9c5890bd371063dc2708cfe18a97675b98493aeb439c29bb941db38 scripts/patches/patch-p0-2-session-once.js
775bb6042e7b5e852a0457eb4a32080822684306e01c3777653f74f436f6bc3b scripts/patches/patch-p0-3-precise-tiering.js
b6c1f36bee8dbad522999a7de93583963078ac7ebff7f404fc5f3d26f1da2b03 scripts/patches/patch-p0-3-reapply-tiering.js b6c1f36bee8dbad522999a7de93583963078ac7ebff7f404fc5f3d26f1da2b03 scripts/patches/patch-p0-3-reapply-tiering.js
96414c5d92dacf9a3f74879fb9e9b2834af92bd6e7d910f079533fd3878735c4 scripts/patches/patch-p0-3-skill-tiering.js 96414c5d92dacf9a3f74879fb9e9b2834af92bd6e7d910f079533fd3878735c4 scripts/patches/patch-p0-3-skill-tiering.js
8b28bf8b0ad9fe914e989d08916552a92b6e4d0870d62b7dcccf829f9c25c855 scripts/patches/patch-p0v2-stop-parallel.js 8b28bf8b0ad9fe914e989d08916552a92b6e4d0870d62b7dcccf829f9c25c855 scripts/patches/patch-p0v2-stop-parallel.js
@ -318,12 +325,20 @@ b4cab8230cc74ec47b5d1a75fd3d5c8db6df0725eefa02df3ee79e7b3bb28a7d scripts/patche
4428f91b4ea67795d1a52b6d0d4482ab16eba8bdb01dcdf481bb5bc4c2810f21 scripts/patches/patch-review-report-required.js 4428f91b4ea67795d1a52b6d0d4482ab16eba8bdb01dcdf481bb5bc4c2810f21 scripts/patches/patch-review-report-required.js
fde139ca7a470f2d3f100e9b26d0aca00234d73a9495de140387657b054e7f15 scripts/patches/patch-review-sealed-frame.js fde139ca7a470f2d3f100e9b26d0aca00234d73a9495de140387657b054e7f15 scripts/patches/patch-review-sealed-frame.js
be50c5c7718cfa8ad338aabbefe44c827e4e69dac54762c83b1e008f02e90b14 scripts/patches/patch-route-accuracy-filter.js be50c5c7718cfa8ad338aabbefe44c827e4e69dac54762c83b1e008f02e90b14 scripts/patches/patch-route-accuracy-filter.js
7faed70942c6c0e54d55474131c0c7c25ff2e7594fc95e5badc25e2d543eab3d scripts/patches/patch-route-interceptor-failopen.js
cb03768fd24c2436625e591cd9a6d14459f170526bc8ca490db3d95e73cdac36 scripts/patches/patch-route-precision-10x-batch-a.js
59c1f20916e12d8ffa4cd00d30e98f005c014549a99ce19d2511e489512cc9b0 scripts/patches/patch-route-precision-10x-batch-b1.js
f757e98bcdaeab954add43c9bba85b826bb9e42d8732fd58ac3a97a9c5bb3838 scripts/patches/patch-route-precision-10x-batch-b2.js
c091165b00687e11d9425af35c9da308c58ca580ccb6671a6921b84c64dad13b scripts/patches/patch-route-precision-10x-batch-b3.js
e25033d9ccf394c59f44defae5e95b917027ec594fd02f4a8656cdf9ab4fd2e5 scripts/patches/patch-route-precision-10x-evo-log.js
51297e0691acdfc34b16d6e383cdb1e75ad8fc9afb7c800b4f7663481c02994d scripts/patches/patch-sanitize-v6-17patterns.js 51297e0691acdfc34b16d6e383cdb1e75ad8fc9afb7c800b4f7663481c02994d scripts/patches/patch-sanitize-v6-17patterns.js
399f00b5d69403a5bffd93bf39e751fe5d1bbfa2837cddb6a0f6271e735aa714 scripts/patches/patch-sanitize-v6-fix-replace.js 399f00b5d69403a5bffd93bf39e751fe5d1bbfa2837cddb6a0f6271e735aa714 scripts/patches/patch-sanitize-v6-fix-replace.js
560073f1285ed3ba769c722149904a0f7391c0c6568f284b90514c896ae62667 scripts/patches/patch-sc-hooks-optimize.js 560073f1285ed3ba769c722149904a0f7391c0c6568f284b90514c896ae62667 scripts/patches/patch-sc-hooks-optimize.js
1114ed94bd4d00d167926d8edd87bd0888abe532a993aa951366b6f652ae9976 scripts/patches/patch-sensitive-paths-delivery-pipeline.js 1114ed94bd4d00d167926d8edd87bd0888abe532a993aa951366b6f652ae9976 scripts/patches/patch-sensitive-paths-delivery-pipeline.js
b7438fb9a000362d2dbce439b27e56384589c045a7bf060d2a53bab09fce9f8f scripts/patches/patch-session-continuity-hooks.js b7438fb9a000362d2dbce439b27e56384589c045a7bf060d2a53bab09fce9f8f scripts/patches/patch-session-continuity-hooks.js
d151ca3e259f76307893ab02a1873a94563c7324f7e08ff7afe9fd1b0e05af0c scripts/patches/patch-session-continuity-timeout.js
d9f59b0ed425ef054c9381af69307eadee5c358315fc95cd75799a354593963c scripts/patches/patch-session-start-memory-audit.js d9f59b0ed425ef054c9381af69307eadee5c358315fc95cd75799a354593963c scripts/patches/patch-session-start-memory-audit.js
d398bf2ab260e8b40d41f420edc95f15ccbb765bfee9c8638db67cc93d7bac3a scripts/patches/patch-skill-cleanup-22.js
706e4e31a08a70a95023ef3cab810ec246f89b355975c7dd24f31c232b468fd8 scripts/patches/patch-ssrf-ipv6-rfc1918.js 706e4e31a08a70a95023ef3cab810ec246f89b355975c7dd24f31c232b468fd8 scripts/patches/patch-ssrf-ipv6-rfc1918.js
a2bde8af2e0ef485c3d67d027f6c832080841eed344245566ce2cf97922496d0 scripts/patches/patch-staging-pipeline-gray-activate.js a2bde8af2e0ef485c3d67d027f6c832080841eed344245566ce2cf97922496d0 scripts/patches/patch-staging-pipeline-gray-activate.js
80f561a8ed28b80dd2336c232fa717fbf2c6e0ca1c89183f4cff2956fe79ced9 scripts/patches/patch-stop-dispatcher-24h-dedup.js 80f561a8ed28b80dd2336c232fa717fbf2c6e0ca1c89183f4cff2956fe79ced9 scripts/patches/patch-stop-dispatcher-24h-dedup.js
@ -363,6 +378,7 @@ cd8971043adfb1a0bc3a89999af69e74cf864e8f64e22be660aad4cee2ebc5b4 scripts/patche
a5801d91c9da458a540aab2c097f70a48ff89385b28dba196f131c1a4a501f14 scripts/patches/patch-x13-handoff-atomic-write.js a5801d91c9da458a540aab2c097f70a48ff89385b28dba196f131c1a4a501f14 scripts/patches/patch-x13-handoff-atomic-write.js
ebcc3daeda23a6b945b0bc4d42ef67956973f2c2ed0cc5f193a58178db7c5be2 scripts/patches/scan-credentials.js ebcc3daeda23a6b945b0bc4d42ef67956973f2c2ed0cc5f193a58178db7c5be2 scripts/patches/scan-credentials.js
bd3ac24cfa4535ebf54635cd8444c78ab3cbbef40fcacb9ae42bd3b5f6b433e2 scripts/patches/test-l1b-arbitration.js bd3ac24cfa4535ebf54635cd8444c78ab3cbbef40fcacb9ae42bd3b5f6b433e2 scripts/patches/test-l1b-arbitration.js
5e64297a4992334cc8e44790dace9ddf99208118117eb4fdbb0f34cc520c6d7a scripts/patches/test-route-regression-0427.js
e126d625cd6c4e0767c52c15ed68a103e85127194b1750e3260348c824ed0807 scripts/patches/token-saver-dispatcher-source.js e126d625cd6c4e0767c52c15ed68a103e85127194b1750e3260348c824ed0807 scripts/patches/token-saver-dispatcher-source.js
68740f19c08589fa06bb15372804d96abc090e4f7dd98413989f05c2e0a694aa scripts/patches/v6.6-rc2-01-register-subagent-stop.js 68740f19c08589fa06bb15372804d96abc090e4f7dd98413989f05c2e0a694aa scripts/patches/v6.6-rc2-01-register-subagent-stop.js
8ce0de8d9f6a49bc2c19fd6b84ffc97bbd37e7f714502f44ec68fa1091eb8600 scripts/patches/v6.6-rc2-02-inject-traceid.js 8ce0de8d9f6a49bc2c19fd6b84ffc97bbd37e7f714502f44ec68fa1091eb8600 scripts/patches/v6.6-rc2-02-inject-traceid.js
@ -386,9 +402,9 @@ e429a59c9d8de78684ebf977f89035aee5690510ff324a6a9abc4024ae7f86d9 scripts/qualit
0ae575d9aaaea3c40dab08347cd7483158877130799a2a14cd69f2cfb3e729e4 scripts/rollback-v6.6-rc2.ps1 0ae575d9aaaea3c40dab08347cd7483158877130799a2a14cd69f2cfb3e729e4 scripts/rollback-v6.6-rc2.ps1
ce6fa67c5ef955aa5a3814e1c34adf274015a1da84ebbfc77b4e58d5601f5371 scripts/route-ab-test.js ce6fa67c5ef955aa5a3814e1c34adf274015a1da84ebbfc77b4e58d5601f5371 scripts/route-ab-test.js
54057aec67898dcbef70766363910a7d7636cdf3f2aa17ca942dcca61a2bbcdf scripts/route-analyzer.js 54057aec67898dcbef70766363910a7d7636cdf3f2aa17ca942dcca61a2bbcdf scripts/route-analyzer.js
d4ba80d68dc8cbb9e95a7bb94d154d9a5e2b15df5a70061a01ee9dcd735d1171 scripts/route-engine.js 462a3c4625d6c3abc65b33da42706e92c01c2c5a282b5cd26e4657d9d0f1e40d scripts/route-engine.js
5f98d4631ee2923137d9c82050af7f350eee4de590b6efda1edb3043d61c95da scripts/route-feedback.js 5f98d4631ee2923137d9c82050af7f350eee4de590b6efda1edb3043d61c95da scripts/route-feedback.js
5832ae388bc31c70ff943e8e9233885cee05140ba4f2e920c2c12559a09dfd69 scripts/route-state.js 7ecc1871a4682d9da7213372f6c740927b231700a24352281674f14d7a2cd4e3 scripts/route-state.js
f7b65549eb6caecfe89ab463a0518e4d9abf60a03b8b510733342b0b0461924c scripts/route-telemetry.js f7b65549eb6caecfe89ab463a0518e4d9abf60a03b8b510733342b0b0461924c scripts/route-telemetry.js
0d6166c316de0aa1ffc43fba65f34b3b5d31c6836cea7870f8717fb601a8c65b scripts/sanitize.js 0d6166c316de0aa1ffc43fba65f34b3b5d31c6836cea7870f8717fb601a8c65b scripts/sanitize.js
227234e869ab040f709724f971ddfd9304f9d0f03595b4201bba0f7d9a156f1a scripts/semantic-scorer.js 227234e869ab040f709724f971ddfd9304f9d0f03595b4201bba0f7d9a156f1a scripts/semantic-scorer.js
@ -419,10 +435,10 @@ b17a6ec98a6dd1afbd42903e9f6db80afa27ff0143a28921d9350941db9607fc scripts/weekly
f9b530a7a13dda8c7edcc0b072c63cd3cb629c5330c095cffabc59d04b7a5dcd scripts/weight-store.js f9b530a7a13dda8c7edcc0b072c63cd3cb629c5330c095cffabc59d04b7a5dcd scripts/weight-store.js
318abc88a501e0e3571bca5e5c790696a39544b6af5f38bfd5aebeac5214a024 scripts/workflow-patterns.js 318abc88a501e0e3571bca5e5c790696a39544b6af5f38bfd5aebeac5214a024 scripts/workflow-patterns.js
b57073d6e491e35e660f4baab926603185a30632aff7bd8b26fe23ead4945ea5 settings.local.template.json b57073d6e491e35e660f4baab926603185a30632aff7bd8b26fe23ead4945ea5 settings.local.template.json
b4c66eb376beacde8ffdeb819f573623d5ac4410889608e1967f6486dd6fb632 settings.template.json 4e295cd1c8961d263fd35dc42889c327e543ed79bc3f1c4213742c62d6ea0eb0 settings.template.json
cff42b637af3463bff15cc77e6a5231969632847bdd052200d4294946f7dd898 SKILL-REGISTRY.md 44bf88501a6ffead43a9cbb88ca7e4cb2c4d0862b318d4af1142b45baab68aa3 SKILL-REGISTRY.md
b87b0db89fc3eab1a7a72e2ad91fe98539567df7972fd2b3cd83efd29cace7e9 skills-index-lite.json 7488a6558cb474e417aafca10c0ef313394a2960dc3034a9f88a51d6c1c76ec1 skills-index-lite.json
532d2bff8add2ae39251c52a4174a059c5cee2d66c3e5d1d60c4914759501f34 skills-index.json e70b03ef9cc33ecad62b525bb155a8f61792d912471cabab74dd7bb5cc3735aa skills-index.json
6079bab64ceab2158740c7c85e960d75f365122bf7ff2afc117d96749e4728ae skills/ai-ml-expert/references/cv-guide.md 6079bab64ceab2158740c7c85e960d75f365122bf7ff2afc117d96749e4728ae skills/ai-ml-expert/references/cv-guide.md
5bbf8dc8360fe5eac7a255ae0b2505d56ada55e1a8b243a94f8dd2a23951e309 skills/ai-ml-expert/references/llm-app.md 5bbf8dc8360fe5eac7a255ae0b2505d56ada55e1a8b243a94f8dd2a23951e309 skills/ai-ml-expert/references/llm-app.md
86745e2ec54760fa37c75c26c5b08890ee13833fc05b9850511476267add9e12 skills/ai-ml-expert/references/pytorch-guide.md 86745e2ec54760fa37c75c26c5b08890ee13833fc05b9850511476267add9e12 skills/ai-ml-expert/references/pytorch-guide.md
@ -431,14 +447,6 @@ e84529b5adde98b49da22acd3ddce21816ee389fed69e01c0b0e441ab38edd8f skills/ai-ml-e
53e5c14a4443bb900990bd7ef811c638aeafa59737ef7e30653160328697325b skills/ai-ml-expert/scripts/evaluate.py 53e5c14a4443bb900990bd7ef811c638aeafa59737ef7e30653160328697325b skills/ai-ml-expert/scripts/evaluate.py
9c8f738da45a282819256b018eb2a11ef1bc30efff34247d297216be0b53be55 skills/ai-ml-expert/scripts/train_utils.py 9c8f738da45a282819256b018eb2a11ef1bc30efff34247d297216be0b53be55 skills/ai-ml-expert/scripts/train_utils.py
af78fc616c059f26196f014d1acad4f2a390d040a8c1416ce8770ca18f9583d2 skills/ai-ml-expert/SKILL.md af78fc616c059f26196f014d1acad4f2a390d040a8c1416ce8770ca18f9583d2 skills/ai-ml-expert/SKILL.md
e30b8dc4f56606833d30657ed4a7c927da5aec5ef8749929b3ba34149a695b8f skills/ai-philosophy-expert/references/ethical-frameworks.md
a19cb705746c05417f21cfb549e99334916411ea5951ce93f9be9dece6e63afe skills/ai-philosophy-expert/SKILL.md
1b5da191af49052473656b1fa6fb1b21b31ce26d4e070bf0ea19a9d4e49a0bd4 skills/angular-architect/references/components.md
bc5350de182909ace9c2676abd482bf3b068eedfe146bd482b6a384a5f314327 skills/angular-architect/references/ngrx.md
536dda628feba51eed0f25bfeb9d8e7a90b9b5f91e69e5b6d5159989a3212bbd skills/angular-architect/references/routing.md
9eff5c67006f71e6f16db248857f9dd277c8d63d2def498d7c56c7cf4b1def65 skills/angular-architect/references/rxjs.md
abf8179b5f23d4b328c82b870a36de1c58e1f78cf5b2d7a1e47e6fd58622bb8f skills/angular-architect/references/testing.md
41467221aac17e0fb4cf4e153179fb292be60c4e79e7ad25dacbef712051a084 skills/angular-architect/SKILL.md
e9e22993c1a9c295c9e45b6ef38d2b38e180bda73287e1347981d2d080c0d04c skills/api-designer/references/error-handling.md e9e22993c1a9c295c9e45b6ef38d2b38e180bda73287e1347981d2d080c0d04c skills/api-designer/references/error-handling.md
e6f2ad2920c00549a717e339484971a1d06ca25b33430edb38fe6ec875f76994 skills/api-designer/references/openapi.md e6f2ad2920c00549a717e339484971a1d06ca25b33430edb38fe6ec875f76994 skills/api-designer/references/openapi.md
9e9294a8f37ea7a1919233d4c6653d3f5a869107edbc521cad0c15eae99dcd04 skills/api-designer/references/pagination.md 9e9294a8f37ea7a1919233d4c6653d3f5a869107edbc521cad0c15eae99dcd04 skills/api-designer/references/pagination.md
@ -543,48 +551,20 @@ c55a72be658eb9a892fd2ec1422fd46840a5e18bce7f9733793595c5e920b2db skills/databas
5ba45f10e01beedf02e2c55257515f16ca710f40b463cc4a871a66c2061f49ea skills/debugger-expert/references/common-errors.md 5ba45f10e01beedf02e2c55257515f16ca710f40b463cc4a871a66c2061f49ea skills/debugger-expert/references/common-errors.md
216af12d387db23a0af66dfec210dae0a9ca15fdfb775864172e667ae747401f skills/debugger-expert/references/debugging-playbook.md 216af12d387db23a0af66dfec210dae0a9ca15fdfb775864172e667ae747401f skills/debugger-expert/references/debugging-playbook.md
0d239e44ccbc76e19c27240df9b99cf4471b4fd05565d43ee2d1577c53c4d42b skills/debugger-expert/SKILL.md 0d239e44ccbc76e19c27240df9b99cf4471b4fd05565d43ee2d1577c53c4d42b skills/debugger-expert/SKILL.md
fb324ab54aa1f141a93c5c0fb6bacc34b4017fc1eada7a0a6ca4ce4ac7166d99 skills/design-consultation/SKILL.md
bb30dd3b9fb17c813f02e74a602902ca145e4dc01ed94f09bdd74758663010fb skills/design-consultation/SKILL.md.tmpl
d35818610add86e3ae5d4c63aa0eeb801e20ea9dedc1e80e1669856b1a0924ed skills/design-review/SKILL.md
d70da394b2612fcf8933041a60eed98f37995f9858e0898550966c71c2b23030 skills/design-review/SKILL.md.tmpl
b7eb373c05c45ab9dc352aa4c3ff08775083189da71ee8d202a14780da8da781 skills/designer-expert/SKILL.md
aa203887b21798f5d9d0b170f5c1be59640700f6f9dd8ff4fa028bf581110ca8 skills/developer-expert/SKILL.md aa203887b21798f5d9d0b170f5c1be59640700f6f9dd8ff4fa028bf581110ca8 skills/developer-expert/SKILL.md
d7c4c77bc3dc07ce9e3faf9a4ce5ea2d6e288376beeb164ff0447338e5d2692f skills/devops-expert/SKILL.md d7c4c77bc3dc07ce9e3faf9a4ce5ea2d6e288376beeb164ff0447338e5d2692f skills/devops-expert/SKILL.md
ba6857aeb23658e6a8d191d2231358d09c1480eacee9b8207417914fa9f5cede skills/devsecops-expert/SKILL.md ba6857aeb23658e6a8d191d2231358d09c1480eacee9b8207417914fa9f5cede skills/devsecops-expert/SKILL.md
be23b81e736046ff5e1b5217a8e0b3a67e855082a163700767ea220d400f94ad skills/diagram-as-code-expert/SKILL.md be23b81e736046ff5e1b5217a8e0b3a67e855082a163700767ea220d400f94ad skills/diagram-as-code-expert/SKILL.md
3195aaa2075162cb5fa7f52d07e012431f5f4961d71a1a3f5251796eeebb87f2 skills/document-release/SKILL.md 3195aaa2075162cb5fa7f52d07e012431f5f4961d71a1a3f5251796eeebb87f2 skills/document-release/SKILL.md
72d5778a051415385db0c605109175b294a0bb8f0916b2905f6aa593ad51a83d skills/document-release/SKILL.md.tmpl 72d5778a051415385db0c605109175b294a0bb8f0916b2905f6aa593ad51a83d skills/document-release/SKILL.md.tmpl
d388a50651a46795c2b4a112f95abe63fad779b8df37acfd8746a9cd002dc047 skills/edge-computing-expert/SKILL.md
ff3b0d17b57f8923c79f61103d0b47670575e257193233845a85a532231d9151 skills/email-communicator/SKILL.md ff3b0d17b57f8923c79f61103d0b47670575e257193233845a85a532231d9151 skills/email-communicator/SKILL.md
bcaebd2101eb6cad1afa9d30cf50f705f209ed3a5f098ef26f638b74103f69e6 skills/evolution-tracker/SKILL.md
a87e5324a2fef28468b77c0f8306e42d48d226d219e413b32b23c12b20335c4d skills/finance-advisor/SKILL.md a87e5324a2fef28468b77c0f8306e42d48d226d219e413b32b23c12b20335c4d skills/finance-advisor/SKILL.md
a8f40a8d0d03abaf4e36b274a08ad5522d9f9b56500a5291e5f7015b23051d74 skills/flutter-expert/references/bloc-state.md
c7529ac3399d8244f13d57289f4d3a268480a2a538bd67607768232a9dd906ef skills/flutter-expert/references/gorouter-navigation.md
e0ea5289f6e8e75f2a7e32b81fc37e17d44c8972e9d822723bdc37a48a5884ea skills/flutter-expert/references/performance.md
3bd3302591f123758cc48bc7cce139e9bb1b86039e1c1e34d69dcc322c0a1c57 skills/flutter-expert/references/project-structure.md
ac0c9d178dd32b5761264d0c4b9584a5d8e1796f59aebf6185c642999515a0cb skills/flutter-expert/references/riverpod-state.md
cef23f70cee95cca0968be205992157ecfeed7a7f6aa0291e7a715bb67ce7d98 skills/flutter-expert/references/widget-patterns.md
8b8d90ea220104f5b916f5bcabbc136cd250783bb73cfffc72872edafc14c235 skills/flutter-expert/SKILL.md
931579d39b983786b9fd14827f9d02f45ec9ef5fd2cf1ad2271160d93737657c skills/frontend-design/SKILL.md
a398bdd862435f7a1993c1f4a89e2771de6c712548ceb6dd484bc0b72c151389 skills/frontend-expert/references/nextjs-guide.md a398bdd862435f7a1993c1f4a89e2771de6c712548ceb6dd484bc0b72c151389 skills/frontend-expert/references/nextjs-guide.md
acadfff2be0540902c8e0ed7bce70f555743f3f5a3674779d518666311f5439b skills/frontend-expert/references/react-patterns.md acadfff2be0540902c8e0ed7bce70f555743f3f5a3674779d518666311f5439b skills/frontend-expert/references/react-patterns.md
c7d850e698fda1e83d48d67693f76c159afb6b5c55d5533920b10acea7f4c2dd skills/frontend-expert/references/state-style-guide.md c7d850e698fda1e83d48d67693f76c159afb6b5c55d5533920b10acea7f4c2dd skills/frontend-expert/references/state-style-guide.md
e2a14a3a4b392fb318cde637f57fc4a53dfd9c169f45b62d49462d0031d5f0b4 skills/frontend-expert/SKILL.md e2a14a3a4b392fb318cde637f57fc4a53dfd9c169f45b62d49462d0031d5f0b4 skills/frontend-expert/SKILL.md
4361f9beaee5ad38824ae6ce611e9ddc80e6a9a9f146757a11f7168dfef91360 skills/genesis-engine/SKILL.md 4361f9beaee5ad38824ae6ce611e9ddc80e6a9a9f146757a11f7168dfef91360 skills/genesis-engine/SKILL.md
96ab7da96bf7114ed6d8efb109de841e2bfca8fccfa91683f06f30a97295dc31 skills/git-operation-master/SKILL.md 96ab7da96bf7114ed6d8efb109de841e2bfca8fccfa91683f06f30a97295dc31 skills/git-operation-master/SKILL.md
be0a0a908690e707b002b59a9acca83620479f7ea3ad189d50e093d943c271ad skills/golang-pro/references/concurrency.md
a8612cd07d898a42f2a5f13782e2b29f9a22ea45e03ee28ecd3ae69cbfd6d195 skills/golang-pro/references/generics.md
7be22df33003ff34b7a54a6d3a8f87e0adf3794b256e18221f5d8048715d59f6 skills/golang-pro/references/interfaces.md
57e8fb221478072c58e674041d059dfc8fd6d2c75024476f9afadefdfd8aa5df skills/golang-pro/references/project-structure.md
5be64624ea0f4d50171e28e0327feca21915a9873e306de54a20673a7e2a1df3 skills/golang-pro/references/testing.md
cd309a1318c284c09c1be791340e47d74a07ecca0f04441d3caf7aa132c5bb2f skills/golang-pro/SKILL.md
6ec4bc60b9c78127ea9e019ef292897886a0d8d395bf61d5c22b31b016b6c44a skills/graphql-architect/references/federation.md
a784ba25054175f67bf01f38027c5522cd8631b16b6d76464174e512b8f58a91 skills/graphql-architect/references/migration-from-rest.md
6f34d5adc6a5b9ff068fa652acf93e8ece8629f3db8946676b4bd307b7b1c559 skills/graphql-architect/references/resolvers.md
6c61db1f7a8c9da3b780a3d4e939a237810c9aece0452d5b801aaf58d04c6db9 skills/graphql-architect/references/schema-design.md
5c01cec3837566ae2e8adbf743c9a52be0a842d494f0ae61cccd37677203dd03 skills/graphql-architect/references/security.md
4d3ad5874ab7900432da2a87daa6b02bdfa7895865c5a6b735d3598efc0bc38f skills/graphql-architect/references/subscriptions.md
121ea14464700187e9b9dc5750f40d6e62936c358b2339e375a69680fe7f1469 skills/graphql-architect/SKILL.md
b385efdf8e9bf51521830f4186153264483de197fed81dd1c314efb907f42973 skills/growth-hacker/SKILL.md b385efdf8e9bf51521830f4186153264483de197fed81dd1c314efb907f42973 skills/growth-hacker/SKILL.md
34220860498dcfc896cae208be7d84fa88b78b709d094df78e1c02841d1945d6 skills/gstack/.agents/skills/gstack-benchmark/SKILL.md 34220860498dcfc896cae208be7d84fa88b78b709d094df78e1c02841d1945d6 skills/gstack/.agents/skills/gstack-benchmark/SKILL.md
0adf6c142bf0bba8e8859e719ac54e16484c9746ea74de70e208ad997d4badf2 skills/gstack/.agents/skills/gstack-browse/SKILL.md 0adf6c142bf0bba8e8859e719ac54e16484c9746ea74de70e208ad997d4badf2 skills/gstack/.agents/skills/gstack-browse/SKILL.md
@ -856,12 +836,6 @@ dd294f5612d93241c333f14a160eaf12edeb377784e3994e51b52e38d82b3d83 skills/nextjs-
81f9bedf42d3fcb80950738f22171eaf6b1b9400865212cead5e7e5b00ee8185 skills/nextjs-developer/SKILL.md 81f9bedf42d3fcb80950738f22171eaf6b1b9400865212cead5e7e5b00ee8185 skills/nextjs-developer/SKILL.md
8b02cd1143efe6106ac181a27a5db304128699d42b85025b9a5f801c81c8a978 skills/notification-system-expert/SKILL.md 8b02cd1143efe6106ac181a27a5db304128699d42b85025b9a5f801c81c8a978 skills/notification-system-expert/SKILL.md
0e2a819a2486d628319682c1af7e823cce1e2106faf9b3a670fb6a85a5a42887 skills/performance-expert/SKILL.md 0e2a819a2486d628319682c1af7e823cce1e2106faf9b3a670fb6a85a5a42887 skills/performance-expert/SKILL.md
355e2431393686eef34cba2fef168e20e415bf6696963928030206d2660afb2a skills/plan-ceo-review/SKILL.md
7875c29f0c57a58841775d9a730662cb064e185c02c8b2d99ff09aa55d02606d skills/plan-ceo-review/SKILL.md.tmpl
8a5eac743de3219b1f81609216cb48caf878b2eab586d1b45e9f6fae50a772f0 skills/plan-design-review/SKILL.md
82c85885b03f0fc3eadbd065d5ea0e5c16b948a2f877fa087e9f02c686d2f535 skills/plan-design-review/SKILL.md.tmpl
ee6d722dbb7f0c6c119928528260fd0d6f2b4f8dd74c2eb5a16844c375ece86f skills/plan-eng-review/SKILL.md
400d4161d9081abf010eca04940a2e1c938c816502d9c980fd7c51a8bbbf996c skills/plan-eng-review/SKILL.md.tmpl
833482920f1883f767e629079228c1c3193ee2352f83d70ebf575de9e8c691e9 skills/planning-with-files/SKILL.md 833482920f1883f767e629079228c1c3193ee2352f83d70ebf575de9e8c691e9 skills/planning-with-files/SKILL.md
e86e75028de3ded6b0f863ca7d9ad25ab03c2a9b131b109d21aa81a9e93b6817 skills/pricing-strategist/SKILL.md e86e75028de3ded6b0f863ca7d9ad25ab03c2a9b131b109d21aa81a9e93b6817 skills/pricing-strategist/SKILL.md
0e279562e6b8e9bdbf0bfdf69c569e73e0d22057686d9ce2757baefb715290fc skills/product-manager-expert/assets/prd-template.md 0e279562e6b8e9bdbf0bfdf69c569e73e0d22057686d9ce2757baefb715290fc skills/product-manager-expert/assets/prd-template.md
@ -896,30 +870,14 @@ b6ac7f955818736c3c34986ed0cd6c3f8df1d81dda9bebd899b24209dc439fea skills/review/
02b7fbefdeb68d32115dabdec14e097e9f1c630a1cde3786dd482fae74aea6e5 skills/reviewer-expert/references/refactoring-catalog.md 02b7fbefdeb68d32115dabdec14e097e9f1c630a1cde3786dd482fae74aea6e5 skills/reviewer-expert/references/refactoring-catalog.md
f94e591c417f2aadac75c345883dfcff35b9060a60f6b0773998a7dabf437a9f skills/reviewer-expert/references/review-checklist.md f94e591c417f2aadac75c345883dfcff35b9060a60f6b0773998a7dabf437a9f skills/reviewer-expert/references/review-checklist.md
f0c60fb969cf9e0220d715ac05dbd1cc574e0aa5a17cff41fb08810dafab04a4 skills/reviewer-expert/SKILL.md f0c60fb969cf9e0220d715ac05dbd1cc574e0aa5a17cff41fb08810dafab04a4 skills/reviewer-expert/SKILL.md
29ff8daa8fbc2b2517aba4a037355e9925e20915dad481998cc825ba2647d6a0 skills/rust-engineer/references/async.md
b8281471fb2a0ab3556f5db8d6524f3bc9ce852e881fb50c86a35f237758e849 skills/rust-engineer/references/error-handling.md
a12cb032e2bd0b5a004763b0454edfb3115c73d528f885371517268339e4efda skills/rust-engineer/references/ownership.md
c2da5cc2f35fbdd8daf7676cb97e0aeec7616c958defff243b6179b1cdfb0a98 skills/rust-engineer/references/testing.md
700ae5dcf2fa5409891f2ae1fd21a34ffb64550e4102a10847785c19a9079e3b skills/rust-engineer/references/traits.md
43c8f4f908ecd74ccad7f47647305189dac575cea50955b8bc3b877884f29730 skills/rust-engineer/SKILL.md
5364985a68b7e57f01a031b8cd70425f152917effc356db2b159be8cb0fc9103 skills/sales-consultant/SKILL.md 5364985a68b7e57f01a031b8cd70425f152917effc356db2b159be8cb0fc9103 skills/sales-consultant/SKILL.md
21d2eca232e729e411e3f1fd903c681a1567e464a1048d7ea7066f576fef504a skills/security-expert/references/auth-patterns.md 21d2eca232e729e411e3f1fd903c681a1567e464a1048d7ea7066f576fef504a skills/security-expert/references/auth-patterns.md
30faaab1b968add1af72160f2615af8f2553c8bfdfceca2f66e719014bf986cb skills/security-expert/references/owasp-top10-guide.md 30faaab1b968add1af72160f2615af8f2553c8bfdfceca2f66e719014bf986cb skills/security-expert/references/owasp-top10-guide.md
1bcadd218e517f7e45f97e92e05c4fc7dd7fc183356171014a5573160607695b skills/security-expert/SKILL.md 1bcadd218e517f7e45f97e92e05c4fc7dd7fc183356171014a5573160607695b skills/security-expert/SKILL.md
27edf75a1d3da93a6fe7bdc24c3a3202af89917e3adae1e2a77f2b5121fbaa6e skills/setup-browser-cookies/SKILL.md
42f2e129afcb4b36670d2156a0f82898e99194ef253b7388ac8f498c178d5fea skills/setup-browser-cookies/SKILL.md.tmpl
b3aab60eb7d814c44d7e51d298596736fe9a3c7db33ff245bc2af3cc9258ea93 skills/setup-deploy/SKILL.md
f88242f5a0882bfa128891995e536b71cc6b2d69fb40f1134632f957c7ea345f skills/setup-deploy/SKILL.md.tmpl
7e2ec54e993dae94c21edd99833910008f96febdca1a519e21666cd5c825536c skills/ship/SKILL.md 7e2ec54e993dae94c21edd99833910008f96febdca1a519e21666cd5c825536c skills/ship/SKILL.md
e50081a5373f2cfcd05f35b00ca31e16b08b0d3fddbe73bfb0624c3c4c14dffb skills/ship/SKILL.md.tmpl e50081a5373f2cfcd05f35b00ca31e16b08b0d3fddbe73bfb0624c3c4c14dffb skills/ship/SKILL.md.tmpl
a827f3d8468564a36df4e09a91a4b7463da13fabfa8dafab906b687f50e710bf skills/social-media-manager/SKILL.md a827f3d8468564a36df4e09a91a4b7463da13fabfa8dafab906b687f50e710bf skills/social-media-manager/SKILL.md
7c506a92523f2702e287a1822ca87348849e632931f640b372212d24338cc5f9 skills/sre-expert/SKILL.md 7c506a92523f2702e287a1822ca87348849e632931f640b372212d24338cc5f9 skills/sre-expert/SKILL.md
7868af160b522770bbd91b8a51e224579788dda7461bcbe7ad5f98a4b4aedc28 skills/swift-expert/references/async-concurrency.md
3ffeee11f91167d6fe4342ce2340204e2c77fa18de8e93767e2e91189953e061 skills/swift-expert/references/memory-performance.md
219a6a7d333396a09e7bbfbba76da0fc8e87209dc09dd4c9966a288d67a2af0e skills/swift-expert/references/protocol-oriented.md
25570394aee504fd14f4b01f7dd8445ffef90cc929c38d6721b5b4e908cb98af skills/swift-expert/references/swiftui-patterns.md
b3b64534d8358e82c7f2a86169b7f9ddc373b9ad228992bfd6f921d246819ada skills/swift-expert/references/testing-patterns.md
2f265b585a70e3acc6a61aa96e9bcb2051a071dee80e363d9f7a63b956c76027 skills/swift-expert/SKILL.md
0e04c748fffb41afee98884708b73f0718a53530eb0da1b70924cb93a19d7e0b skills/tech-lead-mentor/SKILL.md 0e04c748fffb41afee98884708b73f0718a53530eb0da1b70924cb93a19d7e0b skills/tech-lead-mentor/SKILL.md
7820df8e7d2b12fce90a812c97f49b87e319b0063d16d0792f76630c812c4872 skills/tech-writer-expert/SKILL.md 7820df8e7d2b12fce90a812c97f49b87e319b0063d16d0792f76630c812c4872 skills/tech-writer-expert/SKILL.md
57164e8ffecb07061db85f090dbd13757cf09e7206005b6d76fa6d4e85cd96dd skills/technical-seo-expert/SKILL.md 57164e8ffecb07061db85f090dbd13757cf09e7206005b6d76fa6d4e85cd96dd skills/technical-seo-expert/SKILL.md
@ -936,44 +894,6 @@ c6b7c3f070ef51561bde003faf2400c8d6898bee3c8346e4f7d052bea667565e skills/typescr
dcf140ae17c30f7f22508711f6ebf5beb2f34945cf4b8ce386e1f31078628bef skills/typescript-pro/references/type-guards.md dcf140ae17c30f7f22508711f6ebf5beb2f34945cf4b8ce386e1f31078628bef skills/typescript-pro/references/type-guards.md
fc18067394ae965f72822c2469a7c059d35b0ea4d4b42cbc8f077e35c0de3d2a skills/typescript-pro/references/utility-types.md fc18067394ae965f72822c2469a7c059d35b0ea4d4b42cbc8f077e35c0de3d2a skills/typescript-pro/references/utility-types.md
190f8b30ed35678cbe5b651dde9d213b167d8e9998e659cd0971bb9511d4a91c skills/typescript-pro/SKILL.md 190f8b30ed35678cbe5b651dde9d213b167d8e9998e659cd0971bb9511d4a91c skills/typescript-pro/SKILL.md
b754c0efffcf54cb889114a9314be1d5af6ceebe65c4c436d713b52018e752c9 skills/ui-ux-pro-max/data/charts.csv
4cc0df2785f340b8bd0ace50bd4068dc13a7c4c6fa4d9d2723271244ce904614 skills/ui-ux-pro-max/data/colors.csv
351fe61ba45b8146c66de6ea7d67297654639d79a94f537b4d0bad570058c28f skills/ui-ux-pro-max/data/icons.csv
010e1e46db271b94e0af6f18973bea86ddba62feeccd62aeeaa6ed40e2ae2db1 skills/ui-ux-pro-max/data/landing.csv
5359e3037dcb84e99c792ea13dd44f5b78ab6a53557afb309cf19140512aaf5b skills/ui-ux-pro-max/data/products.csv
904c8afcda229629545912dde0e8ac37503757131f0169f80b016f1f58c4fd3f skills/ui-ux-pro-max/data/react-performance.csv
ad18dae3ab6d148d37d144592df80dc1825b6c9e86d9d2d68fdda77434206a37 skills/ui-ux-pro-max/data/stacks/astro.csv
e470e6bc2bcd8562667cc764d1e1afb0be0a30ca4dc9ec12fa18b19d46f156bd skills/ui-ux-pro-max/data/stacks/flutter.csv
1f004891dd189f6bc2a7717401e3100244326e7e2f3963a79bf7b2349d85e091 skills/ui-ux-pro-max/data/stacks/html-tailwind.csv
6c8fd4b0391c342c12b0af15610cd5becbeccaef0bbe8dcfc01d928c3195d93e skills/ui-ux-pro-max/data/stacks/jetpack-compose.csv
06a8e8c13bd44696c6ebb4cb6c4d8dea440e3de8cce77616512529ed29258f0c skills/ui-ux-pro-max/data/stacks/nextjs.csv
05d6e74501b2b6a636faed32450622759f49a5bcafad971f5f4e7eba0eb7be71 skills/ui-ux-pro-max/data/stacks/nuxt-ui.csv
f7a3f2d9542856d00199f85c6c023b7c284cc698e2e6e12d61afce2edacb24d6 skills/ui-ux-pro-max/data/stacks/nuxtjs.csv
a08ca77fcf6b6d9531982dce465366296013bfcf12d2938ac72ad57cf0c4f085 skills/ui-ux-pro-max/data/stacks/react-native.csv
b6e3eb0e4c9c01bd00ffb951d5ca224a19d488b4544c75436f48cf92ff12bf45 skills/ui-ux-pro-max/data/stacks/react.csv
395c2e415ef6f48a474acccdc8eb2c7258a6ed751a5037ba171a2d8823e02ded skills/ui-ux-pro-max/data/stacks/shadcn.csv
bc8d158994a8bdf412830a9dd5e5a5534edd2f367a6e50e0ebd4d76f592dfa3f skills/ui-ux-pro-max/data/stacks/svelte.csv
e87cb9eee208bb753bdab96d4a261c7f830c4317ee90ec098316b248fc32be85 skills/ui-ux-pro-max/data/stacks/swiftui.csv
9b793c3f4eec85f77a758c09322f2b62f3b3ca549bb713ccf0849ef885934395 skills/ui-ux-pro-max/data/stacks/vue.csv
ad3c3236134d833bda01a2faf31ddeaaa1dbd5005fa430dec571fa91b26e5fea skills/ui-ux-pro-max/data/styles.csv
64695e05ac335ba534aff396c4ba9241c642f80ea53b424251d2669b71c4772b skills/ui-ux-pro-max/data/typography.csv
58fc52fff69f62709f8c7dd460de6f6693699ea820f24287f3301a427c2b554e skills/ui-ux-pro-max/data/ui-reasoning.csv
1870ee048f2a2bdd60709f8f7adf7f3b6dcad560bc005c8b2915a8ac8639820d skills/ui-ux-pro-max/data/ux-guidelines.csv
10bfc24e09a6a15db8309e6ef9b6881e1d4af97dffb7c99418aefb8e30df7d54 skills/ui-ux-pro-max/data/web-interface.csv
5459d1f04eea03dfb913742d0c03edc5065ac67f6227e4807fa87699662588cb skills/ui-ux-pro-max/scripts/core.py
4da1d341f3c7749df51b51db4a543a48a427c3c746eb0e9882a1ab86acf3bb54 skills/ui-ux-pro-max/scripts/design_system.py
a449e57060ef5134e98e62fd11f6e240153afd68c27dab89cbb72a3ce55f1498 skills/ui-ux-pro-max/scripts/search.py
a56fbc9d048635c9a0953033a168ee0ff7cc672c8edb3663ed71470c4f89806f skills/ui-ux-pro-max/SKILL.md
9828d8da2e1161025327a239cad31063a8cab6f6fef78a6a6d3b69d05d9a5b79 skills/ultimate-code-expert/SKILL.md
c72faf2b168976fb71884f6f6b7f647259170066f072c94c6e65b38b87f12b9e skills/ux-researcher/SKILL.md
e0dbb382c290579394b4604a76452246dae2a84ab93b2cc1337aceaaaede4645 skills/vue-expert/references/build-tooling.md
e0a7d19ad691df2ed69361571c846eb7734327e5b568d98ef3b120c3cb20834a skills/vue-expert/references/components.md
19887431c295a5b43f899a7b8c50223db60c432b7a93be6cf8372621ddb001a6 skills/vue-expert/references/composition-api.md
5f172340800332503597bba971039e282f61a3a1d850367dfafb6007a7e539a0 skills/vue-expert/references/mobile-hybrid.md
d105a370ea54e8dbd53f67359343c10f30196e36c53bcd6f01b8431d954612bd skills/vue-expert/references/nuxt.md
5cb04b1d9945e73c7a298dc0d2e4a2071d5e80faac8188e09a90c548897fbfbb skills/vue-expert/references/state-management.md
6615f9af00f44fa46af52e25307f924d61763f54cbdd18ef78bfafc0431c5b70 skills/vue-expert/references/typescript.md
fa82ba39a70b4c075ad9aafbb77f2d9715b80232efa76208802ff7c093c963fa skills/vue-expert/SKILL.md
ad96d04be020edb9bdf2633bbba69ae231d262dd27abe3abba2f41b9ee12c80e skills/websocket-engineer/references/alternatives.md ad96d04be020edb9bdf2633bbba69ae231d262dd27abe3abba2f41b9ee12c80e skills/websocket-engineer/references/alternatives.md
05c663a2542c579cdc08316c3781ae361af3ad5eea2408e05986f0b29d58e071 skills/websocket-engineer/references/patterns.md 05c663a2542c579cdc08316c3781ae361af3ad5eea2408e05986f0b29d58e071 skills/websocket-engineer/references/patterns.md
0a31fd106d36e01045f274e5021f5b70323ce1ada448338b63233c1f13dd2ff0 skills/websocket-engineer/references/protocol.md 0a31fd106d36e01045f274e5021f5b70323ce1ada448338b63233c1f13dd2ff0 skills/websocket-engineer/references/protocol.md
@ -982,13 +902,13 @@ a24cebb099481c62776172b8465e6b9a5532ffa993cd43da571656c51aa67844 skills/websock
185d176a5220e0c7a242e0eddb3254cfc4f94e85ec492df460cfc29bd75ff034 skills/websocket-engineer/SKILL.md 185d176a5220e0c7a242e0eddb3254cfc4f94e85ec492df460cfc29bd75ff034 skills/websocket-engineer/SKILL.md
97a7a17c4ba3b2f204702d03e855f53f124ba80d66482817f4d337f55b9bd47f skills/workflow-automation-expert/SKILL.md 97a7a17c4ba3b2f204702d03e855f53f124ba80d66482817f4d337f55b9bd47f skills/workflow-automation-expert/SKILL.md
2f9b75cd4aaca7e1ff5aed6c733c28fde144071e08b2a1fc91936a4baa95dcb7 skills/zero-defect-guardian/SKILL.md 2f9b75cd4aaca7e1ff5aed6c733c28fde144071e08b2a1fc91936a4baa95dcb7 skills/zero-defect-guardian/SKILL.md
aa4cf46d2de1ad399f2d940716d1ab89f378e2fa34fdfea9a9c0d14015b5cad5 stats-compiled.json 3e43144a3c297ab3a9075319def7ed4b9ed2e5a56acc6aa49705eb76c77fe139 stats-compiled.json
bd2110109143f19c0ce9bc41f6d03eaabe951b5b9fbde903ea57ac75b0f2ee1e templates/CLAUDE-portable.md bd2110109143f19c0ce9bc41f6d03eaabe951b5b9fbde903ea57ac75b0f2ee1e templates/CLAUDE-portable.md
54e812f25fb7777acb8688c6d04dc5ff1ec6f481ccd8ac24d675feaf469172f3 templates/settings.portable.json 54e812f25fb7777acb8688c6d04dc5ff1ec6f481ccd8ac24d675feaf469172f3 templates/settings.portable.json
2cfc628805538fe80732180f63ef5b2a0670afcc1cfc17269a2f73c51f1a879f tests/browserbase-wrapper-env.test.js 2cfc628805538fe80732180f63ef5b2a0670afcc1cfc17269a2f73c51f1a879f tests/browserbase-wrapper-env.test.js
99f4bc0c7fb2dc3f1dc02a99f57d9b87192c739d89200cc1e2d766da534d1992 tests/v59-regression.test.js 99f4bc0c7fb2dc3f1dc02a99f57d9b87192c739d89200cc1e2d766da534d1992 tests/v59-regression.test.js
b5b75baeeda0052210bf072c1b296dd29402b1bb85fdd45f1d37a460965daeb1 tools/bookworm-sync.ps1 b5b75baeeda0052210bf072c1b296dd29402b1bb85fdd45f1d37a460965daeb1 tools/bookworm-sync.ps1
0dd856fc7f1aeaa11cef712894949ef1eeb3b758436f8d5a2bf120b5bffbb546 tools/export.mjs f4659b0e59b06815d6341bcd970cb000c4f1e72b5df707868e6897450aafe168 tools/export.mjs
ad98b65635d5375e491a39a668cb848a65034e6cbf2ef3e59d05aea6084331aa tools/scrubber.mjs ad98b65635d5375e491a39a668cb848a65034e6cbf2ef3e59d05aea6084331aa tools/scrubber.mjs
2b8b994419b8d22bf2d29d84aa098d047900244ce7eb6036807703c1b3485859 tools/third-machine-install.ps1 2b8b994419b8d22bf2d29d84aa098d047900244ce7eb6036807703c1b3485859 tools/third-machine-install.ps1
82396552835868194c4604eaeb8b3e33be7e243b62a6973d5d8d767568231b10 VERSION 467b5ecd05aef4657c7d9dbec1976d0a377e072262c46a91a60db2637bd1a4cb VERSION

2
INTEGRITY.sha256.sig
View File

@ -1 +1 @@
e2c8091379c81deafedfc6d0a23469fbcab82eaf715336ce38cda60f149f07287b7d2b08dabab9a22f1e2837315306b057d2b54032fb3c99ef85a2cba1e51d0f ae932670583532fe94790fec7e9c94fd04f6dd9e3b4483e8401a7406aca5a615bf4d68e99c565991319f3dfb441fe5d240a07911474d07dbab77f9292b4ee70e

6
MANIFEST.json
View File

@ -1,6 +1,6 @@
{ {
"version": "6.7.0", "version": "6.6.0",
"exportedAt": "2026-04-27T09:55:18.226Z", "exportedAt": "2026-04-27T14:14:30.172Z",
"fileCount": 994, "fileCount": 914,
"pubKeyFingerprint": "26b83e1b38cdf64a" "pubKeyFingerprint": "26b83e1b38cdf64a"
} }

8
SKILL-REGISTRY.md
View File

@ -1,11 +1,11 @@
# Skill Registry — 技能清单 v6.6.0-phase1-B # Skill Registry — 技能清单 v6.6.1
> 唯一信源: 各 `skills/*/SKILL.md` 的 `description` 字段。本文件为索引视图。 > 唯一信源: 各 `skills/*/SKILL.md` 的 `description` 字段。本文件为索引视图。
## 统计 ## 统计
- **总计**: 95 (59 stable + 1 beta + 35 imported, 38 composable, 7 deprecated 不计入总数) - **总计**: 95 (59 stable + 1 beta + 35 imported, 51 composable, 7 deprecated 不计入总数)
- **最后更新**: 2026-04-24 - **最后更新**: 2026-04-27 (v6.6.1)
### Beta 毕业标准 ### Beta 毕业标准
beta → stable 需满足全部条件: beta → stable 需满足全部条件:
@ -345,4 +345,4 @@ beta → stable 需满足全部条件:
--- ---
*最后更新: 2026-04-03 (v6.5.1)* *最后更新: 2026-04-27 (v6.6.1)*

2
VERSION
View File

@ -1 +1 @@
6.7.0 6.6.0

2
agents/code-reviewer.md
View File

@ -2,7 +2,7 @@
name: code-reviewer name: code-reviewer
description: Use this agent when the user needs automated code review before committing or merging, including security audits, performance analysis, code quality checks, type safety verification, and boundary handling validation. This agent performs multi-dimensional read-only analysis and produces structured review reports with severity levels. description: Use this agent when the user needs automated code review before committing or merging, including security audits, performance analysis, code quality checks, type safety verification, and boundary handling validation. This agent performs multi-dimensional read-only analysis and produces structured review reports with severity levels.
allowed-tools: "Read, Glob, Grep, Bash, WebFetch, WebSearch" allowed-tools: "Read, Glob, Grep, Bash, WebFetch, WebSearch"
model: opus model: sonnet
--- ---
## 调用示例 ## 调用示例

2
agents/delivery-quality-assessor.md
View File

@ -19,7 +19,7 @@ description: |
- 竞争对比分析 (vs 原生 Claude Code / Cursor / Copilot) - 竞争对比分析 (vs 原生 Claude Code / Cursor / Copilot)
- 效率量化 (每会话节省时间、减少纠正次数、安全返工避免) - 效率量化 (每会话节省时间、减少纠正次数、安全返工避免)
allowed-tools: "Read, Glob, Grep, Bash, WebFetch, WebSearch" allowed-tools: "Read, Glob, Grep, Bash, WebFetch, WebSearch"
model: opus model: sonnet
--- ---
# 交付质量评估智能体 (Delivery Quality Assessor) # 交付质量评估智能体 (Delivery Quality Assessor)

2
agents/orchestrator.md
View File

@ -46,7 +46,7 @@ description: >
</commentary> </commentary>
</example> </example>
allowed-tools: "Agent, Read, Glob, Grep, Bash, WebFetch, WebSearch" allowed-tools: "Agent, Read, Glob, Grep, Bash, WebFetch, WebSearch"
model: sonnet model: opus
--- ---
# Orchestrator — 多智能体编排中枢 # Orchestrator — 多智能体编排中枢

2
agents/security-hardener.md
View File

@ -23,7 +23,7 @@ description: |
- 不修改业务逻辑，只修复安全层 - 不修改业务逻辑，只修复安全层
- hooks/ 下文件通过补丁脚本修改 (受 block-sensitive-files 保护) - hooks/ 下文件通过补丁脚本修改 (受 block-sensitive-files 保护)
allowed-tools: "Read, Edit, Write, Glob, Grep, Bash, WebFetch, WebSearch" allowed-tools: "Read, Edit, Write, Glob, Grep, Bash, WebFetch, WebSearch"
model: opus model: sonnet
--- ---
# 安全加固修复智能体 (Security Hardener) # 安全加固修复智能体 (Security Hardener)

12
hooks/checksums.json
View File

@ -37,7 +37,7 @@
"rollback-on-fail.js": "fe01f02e0d45b4d5f8890c226a6433d0dcc6f68932463ecbed3df1ba12ac0a09", "rollback-on-fail.js": "fe01f02e0d45b4d5f8890c226a6433d0dcc6f68932463ecbed3df1ba12ac0a09",
"route-auditor.js": "f584ab2984108bcf2ff5179bbb4472e3c4b8b5bb86fcd0c15cb0b82649732323", "route-auditor.js": "f584ab2984108bcf2ff5179bbb4472e3c4b8b5bb86fcd0c15cb0b82649732323",
"route-compliance-gate.js": "8caf07ee3ee4df370e058dbf54a3a91450eb5ba63fe672051ede2e2d29dc30c0", "route-compliance-gate.js": "8caf07ee3ee4df370e058dbf54a3a91450eb5ba63fe672051ede2e2d29dc30c0",
"route-interceptor-bundle.js": "792cb71f4e4379aac7faaf1c00e05f4eb65e1db7052ab7b38d2c47ec02e81db1", "route-interceptor-bundle.js": "b9a88af4576f3dcb04e49fcf849721e8163dae052197b76e9419fece52c38e32",
"security-startup-guard.js": "26084c1218f7b8067caddf33a865f03fc6ca561f26432f24b21f159b69568812", "security-startup-guard.js": "26084c1218f7b8067caddf33a865f03fc6ca561f26432f24b21f159b69568812",
"session-heartbeat.js": "cfc7a941c87a67528268be46d19822e69eb159a566da4bf941cac249a76521f2", "session-heartbeat.js": "cfc7a941c87a67528268be46d19822e69eb159a566da4bf941cac249a76521f2",
"session-start-mcp-probe.js": "93092327041326bf864e2635dfd722631a01e62060906d2c547ec56d3e3cc2a8", "session-start-mcp-probe.js": "93092327041326bf864e2635dfd722631a01e62060906d2c547ec56d3e3cc2a8",
@ -48,7 +48,7 @@
"subagent-route-injector.js": "195e2e58d1fdc33125a18cb024019cce1835efe09d600750523e756416327018", "subagent-route-injector.js": "195e2e58d1fdc33125a18cb024019cce1835efe09d600750523e756416327018",
"suggest-tests.js": "f6efeb7093b69ca7efba710bbb3234c511e521085a22cbaae291fa9779b569c4", "suggest-tests.js": "f6efeb7093b69ca7efba710bbb3234c511e521085a22cbaae291fa9779b569c4",
"token-saver-dispatcher.js": "d1bbf5d21b2efe96572a82f891c4f1f9bca2da6c96066885361ff8837c86ffca", "token-saver-dispatcher.js": "d1bbf5d21b2efe96572a82f891c4f1f9bca2da6c96066885361ff8837c86ffca",
"lib/fail-mode.js": "6ff44a5e8427fde8024152ddb5680093875731fb8da0757ead6c7ba1de989570", "lib/fail-mode.js": "8d7e4508dda23416f454e4e4571b8e0cc76ebae467c43b785647ff072d83c9c9",
"lib/fast-cache.js": "58d6ef4ffa50f69944d43b5551cc6f8adfe84de3166defac1d443a6083957bfc", "lib/fast-cache.js": "58d6ef4ffa50f69944d43b5551cc6f8adfe84de3166defac1d443a6083957bfc",
"lib/jsonl-hmac.js": "0683f0c43f65c80a159ef700f7f975b07f2187df61f08130099bf582a6486c44", "lib/jsonl-hmac.js": "0683f0c43f65c80a159ef700f7f975b07f2187df61f08130099bf582a6486c44",
"lib/metrics.js": "63792b207b6148ddc6bb2a8a2a24e3309cfc3e66b4773d4eeed9d811da3ef997", "lib/metrics.js": "63792b207b6148ddc6bb2a8a2a24e3309cfc3e66b4773d4eeed9d811da3ef997",
@ -85,7 +85,7 @@
"scripts/dashboard.js": "c3a0f0ad8050e9b875fc0aee4deac11f16dd83af7f4350e6934fa0f8d8ea5e3d", "scripts/dashboard.js": "c3a0f0ad8050e9b875fc0aee4deac11f16dd83af7f4350e6934fa0f8d8ea5e3d",
"scripts/deploy-portable.js": "4e5492032e27e3c0d5c3d5e7c943daa175883ebd7ed43d67185a3241420596aa", "scripts/deploy-portable.js": "4e5492032e27e3c0d5c3d5e7c943daa175883ebd7ed43d67185a3241420596aa",
"scripts/deterministic-quality-gate.js": "77abe264191760b2d46e919f4a4fd4e345cb573bbdd8ccd88e654aa74b5dc894", "scripts/deterministic-quality-gate.js": "77abe264191760b2d46e919f4a4fd4e345cb573bbdd8ccd88e654aa74b5dc894",
"scripts/disambiguation-rules.json": "8b75e8f538af92f61371d94672d07ff874441ef91badfc3939e53a515e2892c3", "scripts/disambiguation-rules.json": "18316935296758d4e5d3bcdd466c29019d221b0c37260f2c790ae7ef8e13b9e6",
"scripts/disambiguation-tree.js": "a744e559196f1da6528760e94d4b85b398ab5760ba98b9421a2804ba36bd7fec", "scripts/disambiguation-tree.js": "a744e559196f1da6528760e94d4b85b398ab5760ba98b9421a2804ba36bd7fec",
"scripts/domain-capacity-manager.js": "269556f0e1baf2a7a72cae6ce0d79e42b30d42f2d56c71de0f641920b3a0d339", "scripts/domain-capacity-manager.js": "269556f0e1baf2a7a72cae6ce0d79e42b30d42f2d56c71de0f641920b3a0d339",
"scripts/domain-classifier.js": "98e14334f33dddbea30d112850e04e1778136471ac5dccc2084a61b2730e48d3", "scripts/domain-classifier.js": "98e14334f33dddbea30d112850e04e1778136471ac5dccc2084a61b2730e48d3",
@ -94,7 +94,7 @@
"scripts/feature-flags.js": "7ec0549511b31e0afb72558b374bd3ebeb111158b233f77651f549e2949353b3", "scripts/feature-flags.js": "7ec0549511b31e0afb72558b374bd3ebeb111158b233f77651f549e2949353b3",
"scripts/fusion-weight-learner.js": "088dc672368ef9f59c64a01613653a6ae755f9c698d3a731fa9cc5cda75f79b1", "scripts/fusion-weight-learner.js": "088dc672368ef9f59c64a01613653a6ae755f9c698d3a731fa9cc5cda75f79b1",
"scripts/generate-skill-index.js": "ada60a5af89c652a3cfb7dd2986189a363814bd236bfc645ab600494a178932a", "scripts/generate-skill-index.js": "ada60a5af89c652a3cfb7dd2986189a363814bd236bfc645ab600494a178932a",
"scripts/generate-stats.js": "724f3a221c67b31d044268325c53dd799fdcf95db13d4b9e48d2252389ca81b6", "scripts/generate-stats.js": "7dfec0c1f3b09386ac7a41d1804dee80e108a1d613dc99ccf73935984fa7c7a6",
"scripts/health-check.js": "ab1e7b60def2f01ea83419af1caa9d24e8cc62f142bb377aa9199d5f15e97b88", "scripts/health-check.js": "ab1e7b60def2f01ea83419af1caa9d24e8cc62f142bb377aa9199d5f15e97b88",
"scripts/hook-priority-scheduler.js": "7be805736c54917c19a8249c360a3e17e9c023ad72956a06d67f6f5fb4120a69", "scripts/hook-priority-scheduler.js": "7be805736c54917c19a8249c360a3e17e9c023ad72956a06d67f6f5fb4120a69",
"scripts/hook-stdin.js": "7a6a6d8b620e7ff93d5eb161b6d076cd9b8b17d757d7c352381d48a66cb244e1", "scripts/hook-stdin.js": "7a6a6d8b620e7ff93d5eb161b6d076cd9b8b17d757d7c352381d48a66cb244e1",
@ -116,9 +116,9 @@
"scripts/quality-analyzer.js": "e429a59c9d8de78684ebf977f89035aee5690510ff324a6a9abc4024ae7f86d9", "scripts/quality-analyzer.js": "e429a59c9d8de78684ebf977f89035aee5690510ff324a6a9abc4024ae7f86d9",
"scripts/route-ab-test.js": "ce6fa67c5ef955aa5a3814e1c34adf274015a1da84ebbfc77b4e58d5601f5371", "scripts/route-ab-test.js": "ce6fa67c5ef955aa5a3814e1c34adf274015a1da84ebbfc77b4e58d5601f5371",
"scripts/route-analyzer.js": "54057aec67898dcbef70766363910a7d7636cdf3f2aa17ca942dcca61a2bbcdf", "scripts/route-analyzer.js": "54057aec67898dcbef70766363910a7d7636cdf3f2aa17ca942dcca61a2bbcdf",
"scripts/route-engine.js": "d4ba80d68dc8cbb9e95a7bb94d154d9a5e2b15df5a70061a01ee9dcd735d1171", "scripts/route-engine.js": "462a3c4625d6c3abc65b33da42706e92c01c2c5a282b5cd26e4657d9d0f1e40d",
"scripts/route-feedback.js": "5f98d4631ee2923137d9c82050af7f350eee4de590b6efda1edb3043d61c95da", "scripts/route-feedback.js": "5f98d4631ee2923137d9c82050af7f350eee4de590b6efda1edb3043d61c95da",
"scripts/route-state.js": "5832ae388bc31c70ff943e8e9233885cee05140ba4f2e920c2c12559a09dfd69", "scripts/route-state.js": "7ecc1871a4682d9da7213372f6c740927b231700a24352281674f14d7a2cd4e3",
"scripts/route-telemetry.js": "f7b65549eb6caecfe89ab463a0518e4d9abf60a03b8b510733342b0b0461924c", "scripts/route-telemetry.js": "f7b65549eb6caecfe89ab463a0518e4d9abf60a03b8b510733342b0b0461924c",
"scripts/sanitize.js": "0d6166c316de0aa1ffc43fba65f34b3b5d31c6836cea7870f8717fb601a8c65b", "scripts/sanitize.js": "0d6166c316de0aa1ffc43fba65f34b3b5d31c6836cea7870f8717fb601a8c65b",
"scripts/semantic-scorer.js": "227234e869ab040f709724f971ddfd9304f9d0f03595b4201bba0f7d9a156f1a", "scripts/semantic-scorer.js": "227234e869ab040f709724f971ddfd9304f9d0f03595b4201bba0f7d9a156f1a",

2
hooks/checksums.sig
View File

@ -1 +1 @@
b41c72a13f1886f4bd65bbf27b9c7fe515bcd204d2e10c72dd8b3f10fc6fb0c9 1fe91263f610dff5fd3b24729fd92dce48952803ebe918977562d3326f5f1610

6
hooks/lib/fail-mode.js
View File

@ -1,4 +1,4 @@
'use strict'; 'use strict';
/** /**
* fail-mode.js — fail-open/fail-closed 决策 API (P1-FAIL-MODE-V1) * fail-mode.js — fail-open/fail-closed 决策 API (P1-FAIL-MODE-V1)
* *
@ -8,13 +8,13 @@
* - feature-flags.json 中读取 features['bookworm.security.failClosed'].mode * - feature-flags.json 中读取 features['bookworm.security.failClosed'].mode
* - mode='off' / 不存在: 完全无操作（保留原 fail-open 行为） * - mode='off' / 不存在: 完全无操作（保留原 fail-open 行为）
* - mode='warn': 记录 evolution-log violation 但放行 * - mode='warn': 记录 evolution-log violation 但放行
* - mode='enforce': 调用方应据此 process.exit(1) * - mode='enforce': 调用方应据此 process.exit(2)
* *
* Usage: * Usage:
* const { failModeDecide } = require('./lib/fail-mode.js'); * const { failModeDecide } = require('./lib/fail-mode.js');
* try { ... } catch (e) { * try { ... } catch (e) {
* const action = failModeDecide('security-startup-guard', e); * const action = failModeDecide('security-startup-guard', e);
* if (action === 'reject') process.exit(1); * if (action === 'reject') process.exit(2);
* // else: 原 fail-open 路径 * // else: 原 fail-open 路径
* } * }
*/ */

81
hooks/route-interceptor-bundle.js
View File

@ -21,15 +21,31 @@
const fs = require('fs'); const fs = require('fs');
const path = require('path'); const path = require('path');
const crypto = require('crypto'); const crypto = require('crypto');
const { safeAppendJsonl } = require('./lib/safe-append.js'); let safeAppendJsonl;
try { ({ safeAppendJsonl } = require('./lib/safe-append.js')); } catch { safeAppendJsonl = () => {}; }
const readStdin = require('./lib/read-stdin.js'); let readStdin;
try { readStdin = require('./lib/read-stdin.js'); } catch { readStdin = () => Promise.resolve(''); }
// === P3-1 BUNDLE: preload routing deps === // === P3-1 BUNDLE: preload routing deps (fail-open: 模块缺失时降级为空路由) ===
// Phase 0 宪法合规拆分: 核心逻辑提取到独立模块 let runRouteEngine, loadSkillsIndex, _engineRequire;
const { runRouteEngine, loadSkillsIndex, safeRequire: _engineRequire } = require('../scripts/route-engine.js'); let buildBWRDirective, _EXEMPT;
const { buildBWRDirective, MUST_INVOKE_EXEMPT_INTENTS: _EXEMPT } = require('../scripts/bwr-builder.js'); let _writeRouteState;
const { writeRouteState: _writeRouteState } = require('../scripts/route-state.js'); let _routingReady = true;
try {
({ runRouteEngine, loadSkillsIndex, safeRequire: _engineRequire } = require('../scripts/route-engine.js'));
({ buildBWRDirective, MUST_INVOKE_EXEMPT_INTENTS: _EXEMPT } = require('../scripts/bwr-builder.js'));
({ writeRouteState: _writeRouteState } = require('../scripts/route-state.js'));
} catch (e) {
_routingReady = false;
runRouteEngine = () => ({ skill: null, confidence: 0, candidates: [] });
loadSkillsIndex = () => [];
_engineRequire = () => null;
buildBWRDirective = () => '[BWR:skip] routing modules unavailable';
_EXEMPT = [];
_writeRouteState = () => {};
process.stderr.write('[route-interceptor] WARN: routing modules not found, degraded to skip mode: ' + (e.message || '') + '\n');
}
// H13: 意图分类器立即加载 (每次必用) // H13: 意图分类器立即加载 (每次必用)
const _preloaded = {}; const _preloaded = {};
@ -326,18 +342,24 @@ function main() {
// 继承尝试函数 (simple + 斧二 + 斧四 共用) // 继承尝试函数 (simple + 斧二 + 斧四 共用)
const INHERIT_WINDOW_MS = 5 * 60 * 1000; const INHERIT_WINDOW_MS = 5 * 60 * 1000;
// IMAGE_INHERIT_LAST_VALID_PRIMARY_v1_APPLIED
function tryInherit() { function tryInherit() {
if (!_cachedPrevState) return null; if (!_cachedPrevState) return null;
const prevTs = _cachedPrevState.ts ? new Date(_cachedPrevState.ts).getTime() : 0; const prevTs = _cachedPrevState.ts ? new Date(_cachedPrevState.ts).getTime() : 0;
const elapsed = Date.now() - prevTs; const elapsed = Date.now() - prevTs;
if ( if (elapsed > INHERIT_WINDOW_MS) return null;
elapsed > INHERIT_WINDOW_MS ||
!_cachedPrevState.routing?.primary || // Item 2: 图片继承链修复 — primary='none' 时回退到 lastValidPrimary
_cachedPrevState.routing.primary === 'none' let prevRouting = _cachedPrevState.routing;
) return null; if (!prevRouting) return null;
const prevRouting = _cachedPrevState.routing; let effectivePrimary = prevRouting.primary;
if (!effectivePrimary || effectivePrimary === 'none') {
const lvp = _cachedPrevState.lastValidPrimary || (prevRouting && prevRouting.lastValidPrimary);
if (!lvp || lvp === 'none') return null;
effectivePrimary = lvp;
}
return { return {
primary: prevRouting.primary, primary: effectivePrimary,
candidates: (prevRouting.candidates || []).map(c => ({ candidates: (prevRouting.candidates || []).map(c => ({
...c, ...c,
confidence: Math.round(c.confidence * 0.7 * 100) / 100, confidence: Math.round(c.confidence * 0.7 * 100) / 100,
@ -349,10 +371,26 @@ function main() {
}; };
} }
// CONFIRM_WORDS_FORCE_INHERIT_v1_APPLIED
// Item 9: 确认词强制继承 — 精确短确认词直接 tryInherit()，不走 TF-IDF
const _CONFIRM_WORDS = ['执行', '开始', '继续', '确认', '好的', '行', '可以', 'go', 'yes', 'proceed', 'ok'];
const _promptTrimmed = prompt.trim().toLowerCase();
const _isConfirmWord = _CONFIRM_WORDS.some(w => _promptTrimmed === w) ||
(_promptTrimmed.length <= 4 && _CONFIRM_WORDS.some(w => _promptTrimmed.includes(w)));
if (isImageQuery) { if (isImageQuery) {
// 斧四: 图片查询 → 强制继承上轮，不走 TF-IDF // 斧四: 图片查询 → 强制继承上轮，不走 TF-IDF
routing = tryInherit() || { primary: 'none', candidates: [], confidence: 0, chain: [] }; routing = tryInherit() || { primary: 'none', candidates: [], confidence: 0, chain: [] };
inherited = routing.primary !== 'none'; inherited = routing.primary !== 'none';
} else if (_isConfirmWord) {
// Item 9: 确认词 → 强制继承，使用 lastValidPrimary 机制
const _confirmInherit = tryInherit();
if (_confirmInherit && _confirmInherit.primary && _confirmInherit.primary !== 'none') {
routing = _confirmInherit;
inherited = true;
} else {
routing = { primary: 'none', candidates: [], confidence: 0, chain: [] };
}
} else if (intent.complexity === 'simple') { } else if (intent.complexity === 'simple') {
// simple: 继承上一次路由 (continue/select/confirm + general/explain) // simple: 继承上一次路由 (continue/select/confirm + general/explain)
const inheritResult = tryInherit(); const inheritResult = tryInherit();
@ -420,6 +458,12 @@ function main() {
} catch {} } catch {}
} }
// COLD_START_CAP_REAPPLY_v1
if (routing._coldStartApplied && routing.candidates && routing.candidates.length >= 2) {
const _capGap = routing.candidates[0].confidence - routing.candidates[1].confidence;
if (_capGap < 0.15 && routing.confidence > 0.65) routing.confidence = 0.65;
}
// v5.3: 记录技能使用到会话记忆 // v5.3: 记录技能使用到会话记忆
if (sessionMemory && routing.primary && routing.primary !== 'none') { if (sessionMemory && routing.primary && routing.primary !== 'none') {
try { try {
@ -428,6 +472,15 @@ function main() {
} }
} }
// Item 2: 维护 lastValidPrimary — 供后续 tryInherit() 使用
if (routing.primary && routing.primary !== 'none') {
routing.lastValidPrimary = routing.primary;
} else if (_cachedPrevState) {
const _oldLvp = _cachedPrevState.lastValidPrimary ||
(_cachedPrevState.routing && _cachedPrevState.routing.lastValidPrimary);
if (_oldLvp && _oldLvp !== 'none') routing.lastValidPrimary = _oldLvp;
}
// 写入 route-state // 写入 route-state
writeRouteState(traceId, prompt, intent, routing); writeRouteState(traceId, prompt, intent, routing);
// [P2-1] SHADOW_HAIKU_v1 // [P2-1] SHADOW_HAIKU_v1

2
package.json
View File

@ -1,6 +1,6 @@
{ {
"name": "bookworm-hooks", "name": "bookworm-hooks",
"version": "6.5.0", "version": "6.6.0",
"private": true, "private": true,
"description": "Bookworm Smart Assistant hooks and tests", "description": "Bookworm Smart Assistant hooks and tests",
"scripts": { "scripts": {

66
scripts/disambiguation-rules.json
View File

@ -1,9 +1,9 @@
{ {
"_meta": { "_meta": {
"version": "1.5.2", "version": "1.5.2",
"description": "消歧规则外部化 — v6.5.1 扩展至 83 条，R81-R83 新增路由精准度修复 (bookworm自检→self-auditor, 自动修复→self-healer, 裸字自动 BAE penalty) | v1.5 (2026-04-24): R84-R88 Bookworm 元词路由修复 | L1d (2026-04-25): R84/R86 追加无 bookworm 短词分支 (路由分析/钩子管线/系统自检 等) | L1d (2026-04-25): R84/R86 追加无 bookworm 短词分支 (路由分析/钩子管线/系统自检 等) | v1.5.1 (2026-04-25): R89 路由自愈场景修复 (D1 Q7) | v1.5.2 (2026-04-25): R87 penalty 清理 + R84 L1d 业务前缀扩展", "description": "消歧规则外部化 — v6.5.1 扩展至 93 条，R81-R83 新增路由精准度修复 (bookworm自检→self-auditor, 自动修复→self-healer, 裸字自动 BAE penalty) | v1.5 (2026-04-24): R84-R88 Bookworm 元词路由修复 | L1d (2026-04-25): R84/R86 追加无 bookworm 短词分支 (路由分析/钩子管线/系统自检 等) | L1d (2026-04-25): R84/R86 追加无 bookworm 短词分支 (路由分析/钩子管线/系统自检 等) | v1.5.1 (2026-04-25): R89 路由自愈场景修复 (D1 Q7) | v1.5.2 (2026-04-25): R87 penalty 清理 + R84 L1d 业务前缀扩展",
"generatedFrom": "route-analyzer.js DISAMBIGUATION_RULES (v6.5.1 消歧 83 条)", "generatedFrom": "route-analyzer.js DISAMBIGUATION_RULES (v6.5.1 消歧 83 条)",
"ruleCount": 89, "ruleCount": 93,
"changelog": [ "changelog": [
"R01: 添加 mutual_exclusion 注解，memory leak 场景排除 performance-expert 误触", "R01: 添加 mutual_exclusion 注解，memory leak 场景排除 performance-expert 误触",
"R05: 添加 mutual_exclusion 注解，memory leak 由 R01 优先处理", "R05: 添加 mutual_exclusion 注解，memory leak 由 R01 优先处理",
@ -58,11 +58,19 @@
"L1d: R86 追加 系统自检 无 bookworm 短词分支 (反回归: 边界字符前缀)", "L1d: R86 追加 系统自检 无 bookworm 短词分支 (反回归: 边界字符前缀)",
"R89: 新增 — 路由/规则/计数 自愈 → self-healer (D1 Q7 修复, 与 R84/R86 对称, penalty vue-expert/api-integration-specialist/reviewer-expert)", "R89: 新增 — 路由/规则/计数 自愈 → self-healer (D1 Q7 修复, 与 R84/R86 对称, penalty vue-expert/api-integration-specialist/reviewer-expert)",
"R87: 修正 — 移除 penalty self-auditor (与 R86 boost 语义冲突, 导致查询型 self-auditor 场景误降权到 developer-expert)", "R87: 修正 — 移除 penalty self-auditor (与 R86 boost 语义冲突, 导致查询型 self-auditor 场景误降权到 developer-expert)",
"R84: L1d 扩展 — trigger negative-lookbehind 增加 uniapp/uni-app/taro/svelte/solid/qwik 业务前缀排除, 防\"uniapp 路由消歧\"等误吸到 self-auditor" "R84: L1d 扩展 — trigger negative-lookbehind 增加 uniapp/uni-app/taro/svelte/solid/qwik 业务前缀排除, 防\"uniapp 路由消歧\"等误吸到 self-auditor",
"R27: 移除 bookworm|自检 关键词 (已由 R81-R89 覆盖)",
"R58: 补充 boost: evolution-tracker",
"R90: 新增 — SRE 专属场景 sre-expert (postmortem/on-call/runbook等)",
"R91: 新增 — 变更影响分析 impact-analyst (爆炸半径/依赖分析等)",
"R92: 新增 — Google Sheets 数据分析 data-analyst-expert (preferred_mcp: google-drive)",
"R93: 新增 — 浏览器MCP统一路由 browser-automation-expert (preferred_mcp: playwright)"
], ],
"l1d_implicit_meta_applied": true, "l1d_implicit_meta_applied": true,
"l1d_patched_at": "2026-04-25T02:11:52.010Z", "l1d_patched_at": "2026-04-25T02:11:52.010Z",
"l1d_implicit_meta_applied_v2": true "l1d_implicit_meta_applied_v2": true,
"PATCH_ROUTE_PRECISION_10X_BATCH_A_APPLIED": true,
"patchedAt_batchA": "2026-04-27T13:06:05.369Z"
}, },
"rules": [ "rules": [
{ {
@ -362,7 +370,7 @@
{ {
"id": "R27", "id": "R27",
"note": "系统自检/健康检查 → project-audit-expert (避免误入 debugger/performance)", "note": "系统自检/健康检查 → project-audit-expert (避免误入 debugger/performance)",
"trigger": "系统自检|系统健康|健康检查|自审计|配置检查|一致性检查|self.?audit|health.?check|系统诊断|bookworm|自检", "trigger": "系统自检|系统健康|健康检查|自审计|配置检查|一致性检查|self.?audit|health.?check|系统诊断",
"boost": "project-audit-expert", "boost": "project-audit-expert",
"penalty": [ "penalty": [
"debugger-expert", "debugger-expert",
@ -727,7 +735,8 @@
"retro" "retro"
], ],
"weight": 0.25, "weight": 0.25,
"description": "系统进化追踪用 evolution-tracker, 团队工程周报用 retro" "description": "系统进化追踪用 evolution-tracker, 团队工程周报用 retro",
"boost": "evolution-tracker"
}, },
{ {
"id": "R59", "id": "R59",
@ -1132,6 +1141,51 @@
], ],
"weight": 0.55, "weight": 0.55,
"description": "写侧自愈动词 → self-healer，与 R84/R86 (read→self-auditor) 对称，penalty vue-router 等误触" "description": "写侧自愈动词 → self-healer，与 R84/R86 (read→self-auditor) 对称，penalty vue-router 等误触"
},
{
"id": "R90",
"note": "SRE 专属场景 → sre-expert，从 devops-expert 分离",
"trigger": "sli|slo|sla.*(?:监控|告警)|on.?call|postmortem|error.*budget|toil|事故响应|incident.*response|runbook|alert.*rule",
"boost": "sre-expert",
"penalty": [
"devops-expert"
],
"weight": 0.35,
"note2": "SRE专属场景从devops-expert中分离"
},
{
"id": "R91",
"note": "变更影响分析 → impact-analyst，与架构咨询消歧",
"trigger": "变更影响|影响分析|影响范围|爆炸半径|依赖分析|改.*(?:会|有).*影响|change.*impact|blast.*radius|downstream.*impact|调用链.*分析|谁在.*(?:用|调用)",
"boost": "impact-analyst",
"penalty": [
"architect-expert",
"developer-expert"
],
"weight": 0.35,
"note2": "变更影响分析与架构咨询消歧"
},
{
"id": "R92",
"note": "Google Sheets 数据分析 → data-analyst-expert (MCP: google-drive)",
"trigger": "(?:google\\s*sheets?|谷歌表格).*(?:分析|统计|可视化|透视|图表|数据清洗|pivot)",
"boost": "data-analyst-expert",
"penalty": [
"developer-expert"
],
"weight": 0.3,
"preferred_mcp": "google-drive",
"note2": "Google Sheets数据分析场景从developer-expert分流"
},
{
"id": "R93",
"note": "浏览器MCP统一路由 → browser-automation-expert (playwright为主)",
"trigger": "(?:browser.?mcp|computer.?control|桌面控制).*(?:测试|自动化|操作)",
"boost": "browser-automation-expert",
"penalty": [],
"weight": 0.25,
"preferred_mcp": "playwright",
"note2": "浏览器MCP统一路由: playwright为主, chrome-devtools为辅, browser-mcp/computer-control-mcp为备选"
} }
] ]
} }

2
scripts/generate-stats.js
View File

@ -102,7 +102,7 @@ function scanSkills() {
const content = fs.readFileSync(path.join(skillsDir, dir, 'SKILL.md'), 'utf8'); const content = fs.readFileSync(path.join(skillsDir, dir, 'SKILL.md'), 'utf8');
if (/maturity:\s*stable/.test(content)) stable++; if (/maturity:\s*stable/.test(content)) stable++;
else if (/maturity:\s*beta/.test(content)) beta++; else if (/maturity:\s*beta/.test(content)) beta++;
if (/composable:/.test(content)) composable++; if (/composable:\s*true/.test(content)) composable++; // COMPOSABLE_REGEX_FIX_v1
} }
return { total: dirs.length, stable, beta, composable, dirs }; return { total: dirs.length, stable, beta, composable, dirs };

48
scripts/patches/bump-v6.6.1.js Normal file
View File

@ -0,0 +1,48 @@
#!/usr/bin/env node
// Bump version from v6.6.0-phase1-B to v6.6.1 across all files
const fs = require('fs');
const path = require('path');
const ROOT = path.resolve(__dirname, '../..');
const OLD = 'v6.6.0-phase1-B';
const NEW = 'v6.6.1';
const SENTINEL = '<!-- patch:bump-v6.6.1 -->';
const targets = [
{ file: 'CLAUDE.md', pattern: OLD },
{ file: 'SKILL-REGISTRY.md', pattern: OLD },
];
let changed = 0;
for (const t of targets) {
const fp = path.join(ROOT, t.file);
if (!fs.existsSync(fp)) { console.log(`SKIP ${t.file} (not found)`); continue; }
let content = fs.readFileSync(fp, 'utf8');
if (content.includes(SENTINEL)) { console.log(`SKIP ${t.file} (already patched)`); continue; }
// BOM strip
if (content.charCodeAt(0) === 0xFEFF) content = content.slice(1);
const before = content;
content = content.replaceAll(t.pattern, NEW);
if (content !== before) {
// Backup
fs.copyFileSync(fp, fp + '.bak');
fs.writeFileSync(fp, content, 'utf8');
console.log(`PATCHED ${t.file}: ${OLD} -> ${NEW}`);
changed++;
} else {
console.log(`SKIP ${t.file} (pattern not found)`);
}
}
// Also fix SKILL-REGISTRY footer date if still old
const regPath = path.join(ROOT, 'SKILL-REGISTRY.md');
if (fs.existsSync(regPath)) {
let reg = fs.readFileSync(regPath, 'utf8');
reg = reg.replace(/\*最后更新: 2026-04-27\b/, '*最后更新: 2026-04-27 (v6.6.1)');
fs.writeFileSync(regPath, reg, 'utf8');
}
console.log(`Done. ${changed} files patched.`);

37
scripts/patches/fix-cold-cap-override-0427.js Normal file
View File

@ -0,0 +1,37 @@
#!/usr/bin/env node
// Fix: session-memory boost 覆盖冷启动 confidence cap 的 bug
// route-interceptor-bundle.js:443 用 candidates[0].confidence 覆盖了 route-engine 返回的 capped 值
// 修复: 在 A/B test 块之后重新应用 cap
const fs = require('fs');
const path = require('path');
const SENTINEL = '// COLD_START_CAP_REAPPLY_v1';
const fp = path.join(__dirname, '..', '..', 'hooks', 'route-interceptor-bundle.js');
let code = fs.readFileSync(fp, 'utf8');
if (code.includes(SENTINEL)) { console.log('SKIP: already patched'); process.exit(0); }
// 定位 A/B test 块结束位置: "} catch {}" 后的 "}"
const abMarker = "routing.experiment = experiment;";
const idx = code.indexOf(abMarker);
if (idx === -1) { console.error('FAIL: cannot find A/B test marker'); process.exit(1); }
// 找到 A/B test 的闭合 catch 块
const afterAB = code.indexOf('} catch {}', idx);
if (afterAB === -1) { console.error('FAIL: cannot find A/B catch block'); process.exit(1); }
const closeBrace = code.indexOf('}', afterAB + '} catch {}'.length);
if (closeBrace === -1) { console.error('FAIL: cannot find closing brace'); process.exit(1); }
const patch = `
${SENTINEL}
if (routing._coldStartApplied && routing.candidates && routing.candidates.length >= 2) {
const _capGap = routing.candidates[0].confidence - routing.candidates[1].confidence;
if (_capGap < 0.15 && routing.confidence > 0.65) routing.confidence = 0.65;
}`;
// 在 A/B test 块的闭合 } 之后插入
const insertPos = closeBrace + 1;
fs.copyFileSync(fp, fp + '.bak');
code = code.slice(0, insertPos) + patch + code.slice(insertPos);
fs.writeFileSync(fp, code, 'utf8');
console.log('PATCHED: cold-start cap re-apply after session-memory + A/B test');

22
scripts/patches/fix-composable-regex-0427.js Normal file
View File

@ -0,0 +1,22 @@
#!/usr/bin/env node
// Fix: generate-stats.js composable 计数正则不精确
// /composable:/ 会把 composable: false 也计入
// 修正为 /composable:\s*true/ 精确匹配
const fs = require('fs');
const path = require('path');
const SENTINEL = '// COMPOSABLE_REGEX_FIX_v1';
const fp = path.join(__dirname, '..', 'generate-stats.js');
let code = fs.readFileSync(fp, 'utf8');
if (code.includes(SENTINEL)) { console.log('SKIP: already patched'); process.exit(0); }
const old = "if (/composable:/.test(content)) composable++;";
const idx = code.indexOf(old);
if (idx === -1) { console.error('FAIL: cannot find composable regex line'); process.exit(1); }
fs.copyFileSync(fp, fp + '.bak');
code = code.slice(0, idx) +
`if (/composable:\\s*true/.test(content)) composable++; ${SENTINEL}` +
code.slice(idx + old.length);
fs.writeFileSync(fp, code, 'utf8');
console.log('PATCHED: composable regex now requires explicit true value');

23
scripts/patches/fix-lvp-persist-0427.js Normal file
View File

@ -0,0 +1,23 @@
#!/usr/bin/env node
// Fix: lastValidPrimary 未持久化到 route-state-current.json
// writeRouteState 只保存 6 个字段，lastValidPrimary 被丢弃
// 导致图片继承链在跨 hook 调用时永远无法回退到上一个有效路由
const fs = require('fs');
const path = require('path');
const SENTINEL = '// LVP_PERSIST_FIX_v1';
const fp = path.join(__dirname, '..', 'route-state.js');
let code = fs.readFileSync(fp, 'utf8');
if (code.includes(SENTINEL)) { console.log('SKIP: already patched'); process.exit(0); }
const marker = 'domain: routing.domain || null,';
const idx = code.indexOf(marker);
if (idx === -1) { console.error('FAIL: cannot find domain field in writeRouteState'); process.exit(1); }
const insertPos = idx + marker.length;
const patch = `\n lastValidPrimary: routing.lastValidPrimary || null, ${SENTINEL}`;
fs.copyFileSync(fp, fp + '.bak');
code = code.slice(0, insertPos) + patch + code.slice(insertPos);
fs.writeFileSync(fp, code, 'utf8');
console.log('PATCHED: lastValidPrimary now persisted in route-state-current.json');

57
scripts/patches/fix-w1-w5-audit-0427.js Normal file
View File

@ -0,0 +1,57 @@
#!/usr/bin/env node
// Fix W1: SKILL-REGISTRY markdown closing **
// Fix W2: hooksRegistered 27→29
// Fix W3: note about archived skills (cosmetic, stats source-of-truth is compile script)
// Fix W4: agentsOpus 7→6, agentsSonnet 10→11
// Fix W5: fail-mode.js comment exit(1)→exit(2)
const fs = require('fs');
const path = require('path');
const ROOT = path.resolve(__dirname, '../..');
// W1: Fix SKILL-REGISTRY.md line 8 missing closing **
const regPath = path.join(ROOT, 'SKILL-REGISTRY.md');
let reg = fs.readFileSync(regPath, 'utf8');
const oldLine = '- **最后更新: 2026-04-27 (v6.6.1)';
const newLine = '- **最后更新**: 2026-04-27 (v6.6.1)';
if (reg.includes(oldLine)) {
fs.copyFileSync(regPath, regPath + '.bak');
reg = reg.replace(oldLine, newLine);
fs.writeFileSync(regPath, reg, 'utf8');
console.log('W1 FIXED: SKILL-REGISTRY.md closing ** added');
} else {
console.log('W1 SKIP: pattern not found or already fixed');
}
// W2+W4: Fix stats-compiled.json counts
const statsPath = path.join(ROOT, 'stats-compiled.json');
let raw = fs.readFileSync(statsPath, 'utf8');
if (raw.charCodeAt(0) === 0xFEFF) raw = raw.slice(1);
const stats = JSON.parse(raw);
const s = stats.summary;
let changed = [];
if (s.hooksRegistered !== 29) { s.hooksRegistered = 29; s.hooksUnregistered = 21; changed.push('W2: hooksRegistered→29'); }
if (s.agentsOpus !== 6) { s.agentsOpus = 6; changed.push('W4a: agentsOpus→6'); }
if (s.agentsSonnet !== 11) { s.agentsSonnet = 11; changed.push('W4b: agentsSonnet→11'); }
if (changed.length) {
stats.generated = new Date().toISOString();
fs.writeFileSync(statsPath, JSON.stringify(stats, null, 2), 'utf8');
console.log('W2+W4 FIXED:', changed.join(', '));
} else {
console.log('W2+W4 SKIP: already correct');
}
// W5: Fix fail-mode.js comment exit(1)→exit(2)
const fmPath = path.join(ROOT, 'hooks', 'lib', 'fail-mode.js');
let fm = fs.readFileSync(fmPath, 'utf8');
const oldComment = "mode='enforce': 调用方应据此 process.exit(1)";
const newComment = "mode='enforce': 调用方应据此 process.exit(2)";
if (fm.includes(oldComment)) {
fs.copyFileSync(fmPath, fmPath + '.bak');
fm = fm.replace(oldComment, newComment);
fs.writeFileSync(fmPath, fm, 'utf8');
console.log('W5 FIXED: fail-mode.js comment exit(1)→exit(2)');
} else {
console.log('W5 SKIP: already fixed or pattern not found');
}
console.log('All patches applied.');

33
scripts/patches/migrate-session-continuity-to-local.js Normal file
View File

@ -0,0 +1,33 @@
#!/usr/bin/env node
// 将 settings.json 中 session-continuity-mcp 从 npm-cache 路径迁移到本地 ~/.claude/mcp/
// 幂等: 已迁移则跳过
const fs = require('fs');
const path = require('path');
const SENTINEL = '/.claude/mcp/claude-session-continuity-mcp';
const OLD_PATH = 'C:/Users/leesu/AppData/Local/npm-cache/_npx/41147f6a3b3ef0bb/node_modules/claude-session-continuity-mcp';
const NEW_PATH = 'C:/Users/leesu/.claude/mcp/claude-session-continuity-mcp';
const SETTINGS = path.join(__dirname, '../../settings.json');
const raw = fs.readFileSync(SETTINGS, 'utf8');
if (raw.includes(SENTINEL) && !raw.includes(OLD_PATH)) {
console.log('[SKIP] 已迁移，无需重复操作');
process.exit(0);
}
if (!raw.includes(OLD_PATH)) {
console.log('[SKIP] 未找到旧路径，无需迁移');
process.exit(0);
}
const bak = SETTINGS + '.bak-' + new Date().toISOString().slice(0, 10).replace(/-/g, '');
fs.copyFileSync(SETTINGS, bak);
console.log('[BAK] ' + bak);
const updated = raw.split(OLD_PATH).join(NEW_PATH);
fs.writeFileSync(SETTINGS, updated, 'utf8');
const count = (raw.match(new RegExp(OLD_PATH.replace(/[/]/g, '\\/'), 'g')) || []).length;
console.log(`[OK] 替换 ${count} 处 npm-cache → local 路径`);
console.log('[VERIFY] 新路径: ' + NEW_PATH);

90
scripts/patches/patch-p0-3-precise-tiering.js Normal file
View File

@ -0,0 +1,90 @@
#!/usr/bin/env node
/**
* P0-3 Precise Re-tiering
* Uses maturity field + route frequency + domain relevance for accurate T1/T2/T3
* Replaces previous coarse tiering (T3 only matched 2/34)
* Idempotent: overwrites existing tier_class/criticality values
*/
'use strict';
const fs = require('fs');
const path = require('path');
const TARGET = path.resolve(__dirname, '..', '..', 'skills-index.json');
if (!fs.existsSync(TARGET)) { process.stderr.write('[SKIP] not found\n'); process.exit(0); }
const src = fs.readFileSync(TARGET, 'utf8');
let index;
try { index = JSON.parse(src); } catch { process.stderr.write('[FAIL] parse\n'); process.exit(1); }
if (!index.skills) { process.stderr.write('[FAIL] no skills\n'); process.exit(1); }
fs.writeFileSync(TARGET + '.bak-p03-precise.' + Date.now(), src);
// T1: 21 high-frequency / high-criticality core skills
const T1 = new Set([
'ai-ml-expert', 'architect-expert', 'backend-builder', 'debugger-expert',
'developer-expert', 'devops-expert', 'frontend-expert', 'git-operation-master',
'guardian', 'mobile-expert', 'performance-expert', 'project-audit-expert',
'prompt-optimizer', 'qa', 'review', 'reviewer-expert', 'security-expert',
'technical-seo-expert', 'tester-expert', 'workflow-automation-expert',
'zero-defect-guardian',
]);
// T2: 28 on-demand skills (proven useful or domain-relevant)
const T2 = new Set([
// Stable T2 (13): actively routed or clearly relevant
'api-integration-specialist', 'browser-automation-expert', 'cloud-native-expert',
'data-engineer-expert', 'database-tuning-expert', 'devsecops-expert',
'diagram-as-code-expert', 'handoff', 'mcp-probe', 'mcp-prune',
'miniprogram-expert', 'product-manager-expert', 'regex-shell-wizard',
// Unknown T2 (10): infra/ops/dev tools
'api-designer', 'cloud-architect', 'data-analyst-expert', 'gstack',
'kubernetes-specialist', 'project-coordinator', 'ship', 'sre-expert',
'tech-writer-expert', 'terraform-engineer',
// Imported T2 (5): actively routed or language skills
'codex', 'investigate', 'nextjs-developer', 'python-pro', 'typescript-pro',
]);
// HIGH criticality: security/audit/guardian types
const HIGH_CRIT = new Set([
'security-expert', 'guardian', 'zero-defect-guardian', 'review',
'debugger-expert', 'developer-expert', 'project-audit-expert',
]);
// Load route frequency from available data
const routeFreq = {};
const dataFiles = [
path.join(__dirname, '..', '..', 'debug', 'shadow-route-log.jsonl'),
path.join(__dirname, '..', '..', 'debug', 'metrics-route.jsonl'),
];
for (const f of dataFiles) {
try {
if (!fs.existsSync(f)) continue;
for (const line of fs.readFileSync(f, 'utf8').split('\n').filter(Boolean)) {
try {
const e = JSON.parse(line);
const name = e.p || e.skill;
if (name && name !== 'none') routeFreq[name] = (routeFreq[name] || 0) + 1;
} catch {}
}
} catch {}
}
let t1 = 0, t2 = 0, t3 = 0;
for (const skill of index.skills) {
const name = skill.name || '';
if (T1.has(name)) {
skill.tier_class = 'T1'; t1++;
} else if (T2.has(name)) {
skill.tier_class = 'T2'; t2++;
} else {
skill.tier_class = 'T3'; t3++;
}
skill.criticality = HIGH_CRIT.has(name) ? 'HIGH' : T1.has(name) ? 'MEDIUM' : 'LOW';
skill.callCount30d = routeFreq[name] || 0;
}
const tmp = TARGET + '.tmp.' + process.pid;
fs.writeFileSync(tmp, JSON.stringify(index, null, 2), 'utf8');
fs.renameSync(tmp, TARGET);
process.stderr.write('[DONE] Precise tiering: T1=' + t1 + ' T2=' + t2 + ' T3=' + t3 + '\n');
process.stderr.write('[DATA] Route frequency populated for ' + Object.keys(routeFreq).length + ' skills\n');

85
scripts/patches/patch-route-interceptor-failopen.js Normal file
View File

@ -0,0 +1,85 @@
#!/usr/bin/env node
/**
* Patch: route-interceptor-bundle.js 硬 require → fail-open
* 原因: 旧版部署缺少 scripts/route-engine.js 等文件时,
* 硬 require 导致整个 UserPromptSubmit hook 崩溃 (MODULE_NOT_FOUND)
* 修复: 包裹 try-catch, 模块缺失时降级为 [BWR:skip]
*/
'use strict';
const fs = require('fs');
const path = require('path');
const TARGET = path.join(__dirname, '..', '..', 'hooks', 'route-interceptor-bundle.js');
const SENTINEL = '_routingReady';
if (!fs.existsSync(TARGET)) {
console.log('[patch] route-interceptor-bundle.js 不存在, 跳过');
process.exit(0);
}
let code = fs.readFileSync(TARGET, 'utf8');
if (code.includes(SENTINEL)) {
console.log('[patch] 已包含 fail-open 保护, 跳过');
process.exit(0);
}
// 备份
const bak = TARGET + '.bak.' + Date.now();
fs.writeFileSync(bak, code);
const OLD = `const { safeAppendJsonl } = require('./lib/safe-append.js');
const readStdin = require('./lib/read-stdin.js');
// === P3-1 BUNDLE: preload routing deps ===
// Phase 0 宪法合规拆分: 核心逻辑提取到独立模块
const { runRouteEngine, loadSkillsIndex, safeRequire: _engineRequire } = require('../scripts/route-engine.js');
const { buildBWRDirective, MUST_INVOKE_EXEMPT_INTENTS: _EXEMPT } = require('../scripts/bwr-builder.js');
const { writeRouteState: _writeRouteState } = require('../scripts/route-state.js');`;
const NEW = `let safeAppendJsonl;
try { ({ safeAppendJsonl } = require('./lib/safe-append.js')); } catch { safeAppendJsonl = () => {}; }
let readStdin;
try { readStdin = require('./lib/read-stdin.js'); } catch { readStdin = () => Promise.resolve(''); }
// === P3-1 BUNDLE: preload routing deps (fail-open: 模块缺失时降级为空路由) ===
let runRouteEngine, loadSkillsIndex, _engineRequire;
let buildBWRDirective, _EXEMPT;
let _writeRouteState;
let _routingReady = true;
try {
({ runRouteEngine, loadSkillsIndex, safeRequire: _engineRequire } = require('../scripts/route-engine.js'));
({ buildBWRDirective, MUST_INVOKE_EXEMPT_INTENTS: _EXEMPT } = require('../scripts/bwr-builder.js'));
({ writeRouteState: _writeRouteState } = require('../scripts/route-state.js'));
} catch (e) {
_routingReady = false;
runRouteEngine = () => ({ skill: null, confidence: 0, candidates: [] });
loadSkillsIndex = () => [];
_engineRequire = () => null;
buildBWRDirective = () => '[BWR:skip] routing modules unavailable';
_EXEMPT = [];
_writeRouteState = () => {};
process.stderr.write('[route-interceptor] WARN: routing modules not found, degraded to skip mode: ' + (e.message || '') + '\\n');
}`;
// 规范化换行符再匹配
const codeNorm = code.replace(/\r\n/g, '\n');
const oldNorm = OLD.replace(/\r\n/g, '\n');
if (!codeNorm.includes(oldNorm)) {
console.error('[patch] 找不到目标代码块, 可能已被修改');
process.exit(1);
}
// 保留原文件换行风格
const eol = code.includes('\r\n') ? '\r\n' : '\n';
const newAdapted = NEW.replace(/\n/g, eol);
code = code.replace(OLD.replace(/\n/g, eol), newAdapted);
if (code === fs.readFileSync(TARGET, 'utf8')) {
// fallback: 用规范化后的内容替换
code = codeNorm.replace(oldNorm, NEW);
}
fs.writeFileSync(TARGET, code);
console.log('[patch] route-interceptor-bundle.js → fail-open 保护已注入');
console.log('[patch] 备份:', bak);

178
scripts/patches/patch-route-precision-10x-batch-a.js Normal file
View File

@ -0,0 +1,178 @@
#!/usr/bin/env node
/**
* patch-route-precision-10x-batch-a.js
* 路由精度10项改进 — Batch A: disambiguation-rules.json 变更
* Item 3: R27 移除 bookworm|自检 关键词
* Item 4: 新增 R90 sre-expert boost
* Item 5: 新增 R91 impact-analyst boost
* Item 7: 新增 R92 Google Sheets 数据分析再路由
* Item 8: R58 补充 evolution-tracker boost
* Item 10: 新增 R93 MCP browser consolidation
*
* 安全性: .bak 备份 + sentinel 幂等检查 + UTF-8 无 BOM 写入
*/
'use strict';
const fs = require('fs');
const path = require('path');
// SENTINEL: 防止重复运行
const SENTINEL = 'PATCH_ROUTE_PRECISION_10X_BATCH_A_APPLIED';
const SCRIPTS_DIR = path.join(__dirname, '..');
const RULES_FILE = path.join(SCRIPTS_DIR, 'disambiguation-rules.json');
const BAK_FILE = RULES_FILE + '.bak';
// ── 读取原文件 ─────────────────────────────────────────
if (!fs.existsSync(RULES_FILE)) {
console.error('[ERROR] disambiguation-rules.json not found:', RULES_FILE);
process.exit(1);
}
const raw = fs.readFileSync(RULES_FILE, 'utf8');
const data = JSON.parse(raw);
// ── 幂等检查 ───────────────────────────────────────────
if (data._meta && data._meta[SENTINEL]) {
console.log('[SKIP] Patch already applied (sentinel found). Nothing to do.');
process.exit(0);
}
// 另一种幂等检查: 如果 R90 已存在，也跳过
if (data.rules && data.rules.some(r => r.id === 'R90')) {
console.log('[SKIP] R90 already exists. Patch appears already applied.');
process.exit(0);
}
// ── 备份 ───────────────────────────────────────────────
fs.writeFileSync(BAK_FILE, raw, 'utf8');
console.log('[BAK] Backed up to', BAK_FILE);
// ── Item 3: R27 — 移除 bookworm 和 自检 关键词 ─────────
let r27Modified = false;
for (const rule of data.rules) {
if (rule.id === 'R27') {
const before = rule.trigger;
// 移除整个 pipe-delimited token: bookworm 和 自检
// 策略: 将 trigger 按 | 分割，过滤掉目标词，再重新 join
// 这样避免跨词的正则副作用（如 系统自检|系统健康 → 系统系统健康）
const tokens = rule.trigger.split('|');
const filtered = tokens.filter(tok => tok !== 'bookworm' && tok !== '自检');
let t = filtered.join('|');
// 清理多余的 | 分隔符（防御性）
t = t.replace(/\|{2,}/g, '|').replace(/^\||\|$/g, '');
rule.trigger = t;
r27Modified = (before !== t);
console.log('[ITEM3] R27 trigger before:', before);
console.log('[ITEM3] R27 trigger after :', t);
console.log('[ITEM3]', r27Modified ? 'MODIFIED' : 'NO_CHANGE (keywords not found)');
break;
}
}
// ── Item 8: R58 — 补充 evolution-tracker boost ─────────
let r58Modified = false;
for (const rule of data.rules) {
if (rule.id === 'R58') {
if (!rule.boost) {
rule.boost = 'evolution-tracker';
r58Modified = true;
console.log('[ITEM8] R58 added boost: evolution-tracker');
} else {
console.log('[ITEM8] R58 already has boost:', rule.boost, '— no change');
}
break;
}
}
// ── Items 4/5/7/10: 追加新规则 R90–R93 ─────────────────
const newRules = [
{
id: 'R90',
note: 'SRE 专属场景 → sre-expert，从 devops-expert 分离',
trigger: 'sli|slo|sla.*(?:监控|告警)|on.?call|postmortem|error.*budget|toil|事故响应|incident.*response|runbook|alert.*rule',
boost: 'sre-expert',
penalty: ['devops-expert'],
weight: 0.35,
note2: 'SRE专属场景从devops-expert中分离',
},
{
id: 'R91',
note: '变更影响分析 → impact-analyst，与架构咨询消歧',
trigger: '变更影响|影响分析|影响范围|爆炸半径|依赖分析|改.*(?:会|有).*影响|change.*impact|blast.*radius|downstream.*impact|调用链.*分析|谁在.*(?:用|调用)',
boost: 'impact-analyst',
penalty: ['architect-expert', 'developer-expert'],
weight: 0.35,
note2: '变更影响分析与架构咨询消歧',
},
{
id: 'R92',
note: 'Google Sheets 数据分析 → data-analyst-expert (MCP: google-drive)',
trigger: '(?:google\\s*sheets?|谷歌表格).*(?:分析|统计|可视化|透视|图表|数据清洗|pivot)',
boost: 'data-analyst-expert',
penalty: ['developer-expert'],
weight: 0.30,
preferred_mcp: 'google-drive',
note2: 'Google Sheets数据分析场景从developer-expert分流',
},
{
id: 'R93',
note: '浏览器MCP统一路由 → browser-automation-expert (playwright为主)',
trigger: '(?:browser.?mcp|computer.?control|桌面控制).*(?:测试|自动化|操作)',
boost: 'browser-automation-expert',
penalty: [],
weight: 0.25,
preferred_mcp: 'playwright',
note2: '浏览器MCP统一路由: playwright为主, chrome-devtools为辅, browser-mcp/computer-control-mcp为备选',
},
];
for (const r of newRules) {
data.rules.push(r);
console.log(`[ITEM${r.id === 'R90' ? '4' : r.id === 'R91' ? '5' : r.id === 'R92' ? '7' : '10'}] Added rule ${r.id}`);
}
// ── 更新元数据 ──────────────────────────────────────────
const totalRules = data.rules.length;
data._meta.ruleCount = totalRules;
data._meta[SENTINEL] = true;
data._meta.patchedAt_batchA = new Date().toISOString();
// 更新 changelog
if (!data._meta.changelog) data._meta.changelog = [];
data._meta.changelog.push(
'R27: 移除 bookworm|自检 关键词 (已由 R81-R89 覆盖)',
'R58: 补充 boost: evolution-tracker',
'R90: 新增 — SRE 专属场景 sre-expert (postmortem/on-call/runbook等)',
'R91: 新增 — 变更影响分析 impact-analyst (爆炸半径/依赖分析等)',
'R92: 新增 — Google Sheets 数据分析 data-analyst-expert (preferred_mcp: google-drive)',
'R93: 新增 — 浏览器MCP统一路由 browser-automation-expert (preferred_mcp: playwright)',
);
// 更新 description 中的规则数量引用 (89条 → 实际数)
if (data._meta.description) {
data._meta.description = data._meta.description.replace(/\d+ 条/g, `${totalRules} 条`);
}
// ── 写入 (UTF-8 无 BOM) ────────────────────────────────
const output = JSON.stringify(data, null, 2) + '\n';
fs.writeFileSync(RULES_FILE, output, 'utf8');
console.log('\n[DONE] disambiguation-rules.json updated.');
console.log(` Total rules: ${totalRules}`);
console.log(` R27 modified: ${r27Modified}`);
console.log(` R58 modified: ${r58Modified}`);
console.log(` New rules added: R90, R91, R92, R93`);
// ── JSON 验证 ──────────────────────────────────────────
try {
const verify = JSON.parse(fs.readFileSync(RULES_FILE, 'utf8'));
console.log(`[VERIFY] JSON valid. rules.length=${verify.rules.length}, ruleCount=${verify._meta.ruleCount}`);
} catch (e) {
console.error('[VERIFY ERROR] JSON parse failed:', e.message);
// 回滚
fs.copyFileSync(BAK_FILE, RULES_FILE);
console.error('[ROLLBACK] Restored from .bak');
process.exit(1);
}

79
scripts/patches/patch-route-precision-10x-batch-b1.js Normal file
View File

@ -0,0 +1,79 @@
#!/usr/bin/env node
/**
* patch-route-precision-10x-batch-b1.js
* 路由精度10项改进 — Batch B1: route-engine.js
* Item 1: 冷启动置信度上限 — coldStartApplied=true 且 gap_1_2 < 0.15 时 confidence 上限 0.65
*
* 安全性: .bak 备份 + sentinel 注释检查 + UTF-8 无 BOM 写入
*/
'use strict';
const fs = require('fs');
const path = require('path');
const SENTINEL = '// COLD_START_CONFIDENCE_CAP_v1_APPLIED';
const TARGET = path.join(__dirname, '..', 'route-engine.js');
const BAK = TARGET + '.bak';
if (!fs.existsSync(TARGET)) {
console.error('[ERROR] route-engine.js not found:', TARGET);
process.exit(1);
}
const src = fs.readFileSync(TARGET, 'utf8');
if (src.includes(SENTINEL)) {
console.log('[SKIP] Patch already applied (sentinel found).');
process.exit(0);
}
// 找到短查询置信度上限 patch 注释后面的正确插入点
// 我们需要在 _finalConfidence 计算结束后（短查询cap之后）插入冷启动cap
// 目标: 在 "CONFIDENCE_CAP_SHORT_QUERY_PATCH_2026_04_20" 块之后、
// "ALIAS_RESOLVER_INJECTED" 注释之前插入
const INSERT_AFTER = ` // === ALIAS_RESOLVER_INJECTED_PHASE2_2026_04_25 ===`;
if (!src.includes(INSERT_AFTER)) {
console.error('[ERROR] Anchor "ALIAS_RESOLVER_INJECTED_PHASE2_2026_04_25" not found. Cannot patch safely.');
process.exit(1);
}
const CAP_CODE = `
${SENTINEL}
// 冷启动置信度上限: coldStartApplied=true 且 rank1/rank2 分差 < 0.15 → cap 0.65
// 防止冷启动 boost 后 gap 较小时系统过度自信
if (coldStartApplied && normalized.length >= 2) {
const _n0 = normalized[0] ? (normalized[0].confidence || 0) : 0;
const _n1 = normalized[1] ? (normalized[1].confidence || 0) : 0;
const gap_1_2 = _n0 - _n1;
if (gap_1_2 < 0.15 && _finalConfidence > 0.65) {
_finalConfidence = 0.65;
try {
const _capLog = JSON.stringify({
t: Date.now(), event: 'cold_start_confidence_cap',
gap: Math.round(gap_1_2 * 1000) / 1000,
original: confidence, capped: 0.65,
primary: normalized[0] && normalized[0].name,
}) + '\\n';
fs.appendFileSync(path.join(DEBUG_DIR, 'confidence-cap.log'), _capLog);
} catch {}
}
}
`;
fs.writeFileSync(BAK, src, 'utf8');
console.log('[BAK] Backed up to', BAK);
const patched = src.replace(INSERT_AFTER, CAP_CODE + INSERT_AFTER);
if (patched === src) {
console.error('[ERROR] String replacement produced no change. Aborting.');
process.exit(1);
}
fs.writeFileSync(TARGET, patched, 'utf8');
console.log('[DONE] Item 1: cold-start confidence cap injected into route-engine.js');
console.log(' Sentinel:', SENTINEL);

178
scripts/patches/patch-route-precision-10x-batch-b2.js Normal file
View File

@ -0,0 +1,178 @@
#!/usr/bin/env node
/**
* patch-route-precision-10x-batch-b2.js
* 路由精度10项改进 — Batch B2: route-interceptor-bundle.js
* Item 2: 图片继承链修复 — tryInherit() 处理 primary='none' 时保留 lastValidPrimary
* Item 9: 确认词强制继承 — 短确认词精确匹配时 force tryInherit()
*
* 注意: 文件含 CRLF 行尾，使用正则 \r?\n 匹配，输出保留原文件行尾格式
*/
'use strict';
const fs = require('fs');
const path = require('path');
const SENTINEL_2 = 'IMAGE_INHERIT_LAST_VALID_PRIMARY_v1_APPLIED';
const SENTINEL_9 = 'CONFIRM_WORDS_FORCE_INHERIT_v1_APPLIED';
const TARGET = path.join(__dirname, '..', '..', 'hooks', 'route-interceptor-bundle.js');
const BAK = TARGET + '.bak';
if (!fs.existsSync(TARGET)) {
console.error('[ERROR] route-interceptor-bundle.js not found:', TARGET);
process.exit(1);
}
let src = fs.readFileSync(TARGET, 'utf8');
// Detect line ending used by file (CRLF or LF)
const CRLF = src.includes('\r\n');
const NL = CRLF ? '\r\n' : '\n';
const alreadyItem2 = src.includes(SENTINEL_2);
const alreadyItem9 = src.includes(SENTINEL_9);
if (alreadyItem2 && alreadyItem9) {
console.log('[SKIP] Both Item 2 and Item 9 patches already applied.');
process.exit(0);
}
fs.writeFileSync(BAK, src, 'utf8');
console.log('[BAK] Backed up to', BAK);
// Helper: join lines using the file's native line ending
function L(...lines) {
return lines.join(NL);
}
// ─────────────────────────────────────────────────────────────
// Item 2: 图片继承链修复
// Replace the tryInherit function body using a regex that handles CRLF/LF
// ─────────────────────────────────────────────────────────────
if (!alreadyItem2) {
// Match the tryInherit function — use a regex with \r?\n throughout
// The function spans from "function tryInherit() {" to its closing "}"
// We match the exact known structure
const tryInheritRegex = /function tryInherit\(\) \{\r?\n\s+if \(!_cachedPrevState\) return null;\r?\n\s+const prevTs = _cachedPrevState\.ts \? new Date\(_cachedPrevState\.ts\)\.getTime\(\) : 0;\r?\n\s+const elapsed = Date\.now\(\) - prevTs;\r?\n\s+if \(\r?\n\s+elapsed > INHERIT_WINDOW_MS \|\|\r?\n\s+!_cachedPrevState\.routing\?\.primary \|\|\r?\n\s+_cachedPrevState\.routing\.primary === 'none'\r?\n\s+\) return null;\r?\n\s+const prevRouting = _cachedPrevState\.routing;\r?\n\s+return \{\r?\n\s+primary: prevRouting\.primary,\r?\n\s+candidates: \(prevRouting\.candidates \|\| \[\]\)\.map\(c => \(\{\r?\n\s+\.\.\.c,\r?\n\s+confidence: Math\.round\(c\.confidence \* 0\.7 \* 100\) \/ 100,\r?\n\s+\}\)\),\r?\n\s+confidence: Math\.round\(\(prevRouting\.confidence \|\| 0\) \* 0\.7 \* 100\) \/ 100,\r?\n\s+chain: prevRouting\.chain \|\| \[\],\r?\n\s+\/\/ 宪法 13\.1: 继承路由保留 mustInvoke 标记\r?\n\s+_inheritedMustInvoke: _cachedPrevState\.mustInvoke \|\| false,\r?\n\s+\};\r?\n\s+\}/;
// Build the replacement using the file's native line ending
const ind6 = ' '; // 6 spaces (inner function body)
const ind8 = ' '; // 8 spaces (inside function)
const replacement = L(
`// ${SENTINEL_2}`,
`${ind6}function tryInherit() {`,
`${ind8}if (!_cachedPrevState) return null;`,
`${ind8}const prevTs = _cachedPrevState.ts ? new Date(_cachedPrevState.ts).getTime() : 0;`,
`${ind8}const elapsed = Date.now() - prevTs;`,
`${ind8}if (elapsed > INHERIT_WINDOW_MS) return null;`,
``,
`${ind8}// Item 2: 图片继承链修复 — primary='none' 时回退到 lastValidPrimary`,
`${ind8}let prevRouting = _cachedPrevState.routing;`,
`${ind8}if (!prevRouting) return null;`,
`${ind8}let effectivePrimary = prevRouting.primary;`,
`${ind8}if (!effectivePrimary || effectivePrimary === 'none') {`,
`${ind8} const lvp = _cachedPrevState.lastValidPrimary || (prevRouting && prevRouting.lastValidPrimary);`,
`${ind8} if (!lvp || lvp === 'none') return null;`,
`${ind8} effectivePrimary = lvp;`,
`${ind8}}`,
`${ind8}return {`,
`${ind8} primary: effectivePrimary,`,
`${ind8} candidates: (prevRouting.candidates || []).map(c => ({`,
`${ind8} ...c,`,
`${ind8} confidence: Math.round(c.confidence * 0.7 * 100) / 100,`,
`${ind8} })),`,
`${ind8} confidence: Math.round((prevRouting.confidence || 0) * 0.7 * 100) / 100,`,
`${ind8} chain: prevRouting.chain || [],`,
`${ind8} // 宪法 13.1: 继承路由保留 mustInvoke 标记`,
`${ind8} _inheritedMustInvoke: _cachedPrevState.mustInvoke || false,`,
`${ind8}};`,
`${ind6}}`
);
if (!tryInheritRegex.test(src)) {
console.error('[ERROR Item2] tryInherit regex did not match. Skipping Item 2 tryInherit patch.');
} else {
src = src.replace(tryInheritRegex, replacement);
console.log('[DONE] Item 2a: tryInherit() patched with lastValidPrimary fallback');
}
// Inject lastValidPrimary maintenance before writeRouteState call
// Match: "// 写入 route-state\r?\n writeRouteState(traceId, prompt, intent, routing);"
const writeStateRegex = /\/\/ 写入 route-state\r?\n(\s+)writeRouteState\(traceId, prompt, intent, routing\);/;
const writeStateMatch = writeStateRegex.exec(src);
if (!writeStateMatch) {
console.error('[ERROR Item2] writeRouteState anchor not found. lastValidPrimary maintenance skipped.');
} else {
const ind = writeStateMatch[1]; // actual indentation of the writeRouteState call
const writeReplacement = L(
`// Item 2: 维护 lastValidPrimary — 供后续 tryInherit() 使用`,
`${ind}if (routing.primary && routing.primary !== 'none') {`,
`${ind} routing.lastValidPrimary = routing.primary;`,
`${ind}} else if (_cachedPrevState) {`,
`${ind} const _oldLvp = _cachedPrevState.lastValidPrimary ||`,
`${ind} (_cachedPrevState.routing && _cachedPrevState.routing.lastValidPrimary);`,
`${ind} if (_oldLvp && _oldLvp !== 'none') routing.lastValidPrimary = _oldLvp;`,
`${ind}}`,
``,
`${ind}// 写入 route-state`,
`${ind}writeRouteState(traceId, prompt, intent, routing);`
);
src = src.replace(writeStateRegex, writeReplacement);
console.log('[DONE] Item 2b: lastValidPrimary maintenance injected before writeRouteState');
}
}
// ─────────────────────────────────────────────────────────────
// Item 9: 确认词强制继承
// Insert confirm-word check before the isImageQuery branch
// ─────────────────────────────────────────────────────────────
if (!alreadyItem9) {
// Match "if (isImageQuery) {\r?\n // 斧四..."
const flowRegex = /if \(isImageQuery\) \{\r?\n(\s+)\/\/ 斧四: 图片查询/;
const flowMatch = flowRegex.exec(src);
if (!flowMatch) {
console.error('[ERROR Item9] isImageQuery flow anchor not found. Skipping Item 9 patch.');
} else {
const innerInd = flowMatch[1]; // indentation inside the if block
const outerInd = innerInd.slice(0, innerInd.length - 2); // one level out (2 spaces less)
const confirmBlock = L(
`// ${SENTINEL_9}`,
`${outerInd}// Item 9: 确认词强制继承 — 精确短确认词直接 tryInherit()，不走 TF-IDF`,
`${outerInd}const _CONFIRM_WORDS = ['执行', '开始', '继续', '确认', '好的', '行', '可以', 'go', 'yes', 'proceed', 'ok'];`,
`${outerInd}const _promptTrimmed = prompt.trim().toLowerCase();`,
`${outerInd}const _isConfirmWord = _CONFIRM_WORDS.some(w => _promptTrimmed === w) ||`,
`${outerInd} (_promptTrimmed.length <= 4 && _CONFIRM_WORDS.some(w => _promptTrimmed.includes(w)));`,
``,
`${outerInd}if (isImageQuery) {`,
`${innerInd}// 斧四: 图片查询`
);
src = src.replace(flowRegex, confirmBlock);
// Now insert the _isConfirmWord branch AFTER the isImageQuery block close
// Find "} else if (intent.complexity === 'simple') {" and prepend the confirm branch
const simpleRegex = /(\} else if \(intent\.complexity === 'simple'\) \{)/;
if (!simpleRegex.test(src)) {
console.error('[ERROR Item9] simple-complexity anchor not found. Confirm branch not injected.');
} else {
const confirmBranch = L(
`} else if (_isConfirmWord) {`,
`${innerInd}// Item 9: 确认词 → 强制继承，使用 lastValidPrimary 机制`,
`${innerInd}const _confirmInherit = tryInherit();`,
`${innerInd}if (_confirmInherit && _confirmInherit.primary && _confirmInherit.primary !== 'none') {`,
`${innerInd} routing = _confirmInherit;`,
`${innerInd} inherited = true;`,
`${innerInd}} else {`,
`${innerInd} routing = { primary: 'none', candidates: [], confidence: 0, chain: [] };`,
`${innerInd}}`,
`${outerInd}} else if (intent.complexity === 'simple') {`
);
src = src.replace(simpleRegex, confirmBranch);
console.log('[DONE] Item 9: confirm-words force-inherit injected into routing flow');
}
}
}
// ── 写入 ──────────────────────────────────────────────────────
fs.writeFileSync(TARGET, src, 'utf8');
console.log('\n[SUMMARY] route-interceptor-bundle.js patched.');
console.log(' Item 2 (image inherit chain):', alreadyItem2 ? 'SKIPPED (already applied)' : 'DONE');
console.log(' Item 9 (confirm words inherit):', alreadyItem9 ? 'SKIPPED (already applied)' : 'DONE');

114
scripts/patches/patch-route-precision-10x-batch-b3.js Normal file
View File

@ -0,0 +1,114 @@
#!/usr/bin/env node
/**
* patch-route-precision-10x-batch-b3.js
* 路由精度10项改进 — Batch B3: fusion-weight-learner 激活诊断 + CLAUDE.md 规则数更新
* Item 6: fusion-weight-learner 激活状态核查
* 诊断结论: learner 已在 stop-dispatcher.js Batch2 中被调用 (race 'implicit→fwl')
* fusion-weights.json 中 corrections=0 是 bootstrap 元数据，不代表 learner 未运行
* root cause: 所有 route-feedback.jsonl 条目均为 routedTo===correctedTo (timeout-confirm)
* 修复: 在 learner 跳过时写入诊断日志，便于后续追踪
* CLAUDE.md: 将 "89条" 更新为实际规则数
*
* 安全性: .bak 备份 + sentinel 检查 + UTF-8 无 BOM 写入
*/
'use strict';
const fs = require('fs');
const path = require('path');
const CLAUDE_ROOT = path.join(require('os').homedir(), '.claude');
const SCRIPTS_DIR = path.join(CLAUDE_ROOT, 'scripts');
const HOOKS_DIR = path.join(CLAUDE_ROOT, 'hooks');
const DEBUG_DIR = path.join(CLAUDE_ROOT, 'debug');
const CLAUDE_MD = path.join(CLAUDE_ROOT, 'CLAUDE.md');
const RULES_FILE = path.join(SCRIPTS_DIR, 'disambiguation-rules.json');
// ── Item 6: 核查 fusion-weight-learner 调用链 ──────────────────
console.log('\n[ITEM6] Verifying fusion-weight-learner activation chain...');
// 1. 检查 stop-dispatcher.js 是否调用了 fusion-weight-learner
const STOP_DISPATCHER = path.join(HOOKS_DIR, 'stop-dispatcher.js');
let item6Status = 'UNKNOWN';
let item6Detail = '';
if (fs.existsSync(STOP_DISPATCHER)) {
const sdSrc = fs.readFileSync(STOP_DISPATCHER, 'utf8');
if (sdSrc.includes('fusion-weight-learner.js')) {
item6Status = 'ALREADY_ACTIVE';
item6Detail = 'fusion-weight-learner.js is called in stop-dispatcher.js Batch2 (race "implicit→fwl")';
console.log('[ITEM6] SKIP — learner already wired in stop-dispatcher.js');
console.log('[ITEM6] Root cause of corrections=0: all route-feedback.jsonl entries have routedTo===correctedTo');
console.log('[ITEM6] Learner runs but returns {status:"skip", reason:"纠正不足2条"} — this is correct behavior');
console.log('[ITEM6] No code change needed. Writing diagnostic note to debug log.');
} else {
item6Status = 'NOT_WIRED';
item6Detail = 'fusion-weight-learner.js NOT found in stop-dispatcher.js — needs wiring';
console.log('[ITEM6] WARN — learner not found in stop-dispatcher. Manual investigation required.');
}
} else {
item6Status = 'STOP_DISPATCHER_MISSING';
item6Detail = 'stop-dispatcher.js not found';
console.log('[ITEM6] ERROR — stop-dispatcher.js not found at', STOP_DISPATCHER);
}
// 2. 写入诊断日志
try {
if (!fs.existsSync(DEBUG_DIR)) fs.mkdirSync(DEBUG_DIR, { recursive: true });
const diagEntry = JSON.stringify({
ts: new Date().toISOString(),
event: 'fusion-weight-learner-audit',
status: item6Status,
detail: item6Detail,
fusionWeightsFile: path.join(DEBUG_DIR, 'fusion-weights.json'),
action: item6Status === 'ALREADY_ACTIVE' ? 'no-change-needed' : 'manual-review-required',
}) + '\n';
fs.appendFileSync(path.join(DEBUG_DIR, 'route-engine-audit.log'), diagEntry);
console.log('[ITEM6] Diagnostic written to debug/route-engine-audit.log');
} catch (e) {
console.error('[ITEM6] Could not write diagnostic log:', e.message);
}
// ── CLAUDE.md 规则数更新 ──────────────────────────────────────
console.log('\n[CLAUDE.MD] Updating disambiguation rule count...');
if (!fs.existsSync(CLAUDE_MD)) {
console.error('[CLAUDE.MD] CLAUDE.md not found:', CLAUDE_MD);
} else {
// 读取实际规则数
let actualRuleCount = null;
try {
const rules = JSON.parse(fs.readFileSync(RULES_FILE, 'utf8'));
actualRuleCount = rules.rules ? rules.rules.length : null;
} catch (e) {
console.error('[CLAUDE.MD] Could not read rules file:', e.message);
}
if (actualRuleCount !== null) {
const mdSrc = fs.readFileSync(CLAUDE_MD, 'utf8');
const BAK_MD = CLAUDE_MD + '.bak';
// 查找并替换规则数引用 (格式: "完整 N 条见 scripts/disambiguation-rules.json")
const ruleRefPattern = /完整\s+\d+\s*条见\s+scripts\/disambiguation-rules\.json/g;
const matches = mdSrc.match(ruleRefPattern);
if (!matches) {
console.log('[CLAUDE.MD] No rule count reference found matching pattern. Skipping.');
} else {
fs.writeFileSync(BAK_MD, mdSrc, 'utf8');
const patched = mdSrc.replace(
ruleRefPattern,
`完整 ${actualRuleCount} 条见 scripts/disambiguation-rules.json`
);
if (patched === mdSrc) {
console.log('[CLAUDE.MD] Content unchanged (already up to date).');
} else {
fs.writeFileSync(CLAUDE_MD, patched, 'utf8');
console.log(`[CLAUDE.MD] Updated rule count to ${actualRuleCount} (was: ${matches[0]})`);
}
}
}
}
console.log('\n[DONE] Batch B3 complete.');
console.log(' Item 6 status:', item6Status);

37
scripts/patches/patch-route-precision-10x-evo-log.js Normal file
View File

@ -0,0 +1,37 @@
#!/usr/bin/env node
/**
* patch-route-precision-10x-evo-log.js
* 路由精度10项改进 — 追加 evolution-log 条目 (seq 120)
*/
'use strict';
const fs = require('fs');
const path = require('path');
const CLAUDE_ROOT = path.join(require('os').homedir(), '.claude');
const DEBUG_DIR = path.join(CLAUDE_ROOT, 'debug');
const EVO_LOG = path.join(DEBUG_DIR, 'evolution-log.jsonl');
const ENTRY = {
seq: 120,
ts: '2026-04-27',
version: 'v6.6.1',
trigger: 'route-precision-10x',
summary: '路由精度10项改进: 置信度上限/图片继承链/R27去bookworm/R90-R93新增/R58补boost/确认词继承/fusion-learner激活',
fix_count: 10,
tags: ['route-engine', 'disambiguation', 'confidence', 'inheritance', 'fusion-weights'],
};
// 幂等: 检查 seq=120 是否已存在
if (fs.existsSync(EVO_LOG)) {
const lines = fs.readFileSync(EVO_LOG, 'utf8').split('\n').filter(Boolean);
if (lines.some(l => { try { return JSON.parse(l).seq === 120; } catch { return false; } })) {
console.log('[SKIP] seq=120 already exists in evolution-log.jsonl');
process.exit(0);
}
}
if (!fs.existsSync(DEBUG_DIR)) fs.mkdirSync(DEBUG_DIR, { recursive: true });
fs.appendFileSync(EVO_LOG, JSON.stringify(ENTRY) + '\n', 'utf8');
console.log('[DONE] Appended seq=120 to evolution-log.jsonl');

38
scripts/patches/patch-session-continuity-timeout.js Normal file
View File

@ -0,0 +1,38 @@
#!/usr/bin/env node
// 幂等补丁: 为 session-continuity-mcp 的 3 个 hook 添加 timeout: 5000
const fs = require('fs');
const path = require('path');
const SETTINGS = path.join(process.env.HOME || process.env.USERPROFILE, '.claude', 'settings.json');
const SENTINEL = '__patch_session_continuity_timeout_v1';
const BAK = SETTINGS + '.bak-sct-' + Date.now();
const raw = fs.readFileSync(SETTINGS, 'utf8');
const cfg = JSON.parse(raw);
if (cfg[SENTINEL]) {
console.log('[SKIP] 补丁已应用');
process.exit(0);
}
fs.copyFileSync(SETTINGS, BAK);
console.log('[BAK]', BAK);
const TARGET = 'claude-session-continuity-mcp/dist/hooks/';
let patched = 0;
for (const [event, groups] of Object.entries(cfg.hooks || {})) {
for (const group of groups) {
for (const hook of (group.hooks || [])) {
if (hook.command && hook.command.includes(TARGET) && !hook.timeout) {
hook.timeout = 5000;
patched++;
console.log(`[PATCH] ${event}: +timeout 5000`);
}
}
}
}
cfg[SENTINEL] = new Date().toISOString();
fs.writeFileSync(SETTINGS, JSON.stringify(cfg, null, 2), 'utf8');
console.log(`[DONE] ${patched} hooks patched`);

68
scripts/patches/patch-skill-cleanup-22.js Normal file
View File

@ -0,0 +1,68 @@
#!/usr/bin/env node
/**
* Skill Cleanup: Remove 22 low-value skills
* - 14 delete (irrelevant/niche/overlapping)
* - 8 merge-remove (redundant groups)
* Moves skill dirs to skills/_archived/, removes from skills-index.json
*/
'use strict';
const fs = require('fs');
const path = require('path');
const CLAUDE_ROOT = path.resolve(__dirname, '..', '..');
const INDEX = path.join(CLAUDE_ROOT, 'skills-index.json');
const SKILLS_DIR = path.join(CLAUDE_ROOT, 'skills');
const ARCHIVE_DIR = path.join(SKILLS_DIR, '_archived');
const TO_REMOVE = [
// Cat 1: language/framework mismatch (6)
'angular-architect', 'flutter-expert', 'golang-pro',
'rust-engineer', 'swift-expert', 'vue-expert',
// Cat 1: extremely niche (4)
'ai-philosophy-expert', 'edge-computing-expert', 'graphql-architect', 'evolution-tracker',
// Cat 1: overlapping with T1 (4)
'ultimate-code-expert', 'frontend-design', 'design-consultation', 'design-review',
// Cat 2: plan variants → planning-with-files covers (3)
'plan-ceo-review', 'plan-design-review', 'plan-eng-review',
// Cat 2: design duplicates → frontend-expert T1 covers (3)
'designer-expert', 'ui-ux-pro-max', 'ux-researcher',
// Cat 2: setup tools → devops-expert covers (2)
'setup-browser-cookies', 'setup-deploy',
];
if (!fs.existsSync(INDEX)) { process.stderr.write('[FAIL] skills-index.json not found\n'); process.exit(1); }
const src = fs.readFileSync(INDEX, 'utf8');
let index;
try { index = JSON.parse(src); } catch { process.stderr.write('[FAIL] parse\n'); process.exit(1); }
fs.writeFileSync(INDEX + '.bak-cleanup.' + Date.now(), src);
const removeSet = new Set(TO_REMOVE);
const before = index.skills.length;
index.skills = index.skills.filter(s => !removeSet.has(s.name));
const removed = before - index.skills.length;
// Write updated index
const tmp = INDEX + '.tmp.' + process.pid;
fs.writeFileSync(tmp, JSON.stringify(index, null, 2), 'utf8');
fs.renameSync(tmp, INDEX);
process.stderr.write('[INDEX] Removed ' + removed + '/' + TO_REMOVE.length + ' from skills-index.json (' + before + ' -> ' + index.skills.length + ')\n');
// Move skill directories to _archived
if (!fs.existsSync(ARCHIVE_DIR)) fs.mkdirSync(ARCHIVE_DIR, { recursive: true });
let moved = 0;
for (const name of TO_REMOVE) {
const src = path.join(SKILLS_DIR, name);
const dst = path.join(ARCHIVE_DIR, name);
if (fs.existsSync(src) && !fs.existsSync(dst)) {
try { fs.renameSync(src, dst); moved++; } catch (e) {
process.stderr.write('[WARN] Failed to move ' + name + ': ' + e.message + '\n');
}
}
}
process.stderr.write('[FILES] Moved ' + moved + ' skill dirs to _archived/\n');
// Summary
const tiers = {};
for (const s of index.skills) tiers[s.tier_class] = (tiers[s.tier_class] || 0) + 1;
process.stderr.write('[DONE] Final: ' + index.skills.length + ' skills (T1=' + (tiers.T1||0) + ' T2=' + (tiers.T2||0) + ' T3=' + (tiers.T3||0) + ')\n');

157
scripts/patches/test-route-regression-0427.js Normal file
View File

@ -0,0 +1,157 @@
#!/usr/bin/env node
'use strict';
/**
* v6.6.1 路由精度回归测试 — 5 测试用例
* 直接调用 route-engine + intent-classifier + disambiguation 验证
*/
const path = require('path');
const fs = require('fs');
const ROOT = path.join(__dirname, '..', '..');
// 加载核心模块
const routeEngine = require(path.join(ROOT, 'scripts', 'route-engine.js'));
const intentClassifier = require(path.join(ROOT, 'scripts', 'intent-classifier.js'));
const cwd = process.cwd();
let passed = 0, failed = 0;
function test(name, prompt, expectPrimary, opts = {}) {
const intent = intentClassifier.classify ? intentClassifier.classify(prompt) : { intents: [], entities: [], modifiers: [], complexity: 'medium' };
const result = routeEngine.runRouteEngine(prompt, cwd, intent);
const primary = result.primary;
const confidence = result.confidence;
const candidates = (result.candidates || []).slice(0, 5);
const coldStart = result._coldStartApplied || false;
const firedRules = (result._firedRules || []).map(r => r.id || r.rule || '').filter(Boolean);
// 检查是否命中期望 skill (主路由或 top-3 候选)
const top3Names = candidates.slice(0, 3).map(c => c.name);
const isPrimaryHit = primary === expectPrimary;
const isTop3Hit = top3Names.includes(expectPrimary);
const hit = isPrimaryHit || (opts.allowTop3 && isTop3Hit);
const status = hit ? 'PASS' : 'FAIL';
if (hit) passed++; else failed++;
console.log(`\n[${status}] ${name}`);
console.log(` prompt: "${prompt}"`);
console.log(` expect: ${expectPrimary}`);
console.log(` got: ${primary} (cf: ${confidence})`);
console.log(` top-3: ${top3Names.join(', ')}`);
console.log(` rules: ${firedRules.length > 0 ? firedRules.join(', ') : '(none)'}`);
console.log(` coldStart: ${coldStart}`);
if (opts.checkCap && coldStart) {
const capApplied = confidence <= 0.65;
console.log(` cap@0.65: ${capApplied ? 'YES' : 'NO (BUG!)'}`);
}
if (!hit) {
console.log(` ** MISMATCH: expected ${expectPrimary}, got ${primary}`);
if (isTop3Hit) console.log(` ** (但在 top-3 候选中)`);
}
}
console.log('=== Bookworm v6.6.1 Route Regression Test ===\n');
// TC1: R90 sre-expert
test('TC1: SLI 监控告警 → sre-expert (R90)',
'SLI 监控告警配置',
'sre-expert');
// TC2: R91 impact-analyst
test('TC2: 函数影响分析 → impact-analyst (R91)',
'改这个函数会影响哪些模块',
'impact-analyst');
// TC3: R92 data-analyst-expert
test('TC3: Google Sheets 数据分析 → data-analyst-expert (R92)',
'从 Google Sheets 分析销售数据',
'data-analyst-expert');
// TC4: 确认词 "执行" — 路由引擎层面应该是低置信度/none (继承在 bundle 层处理)
// 这里验证路由引擎不会错误地高置信度命中无关 skill
test('TC4: 确认词 "执行" (路由引擎层)',
'执行',
'none',
{ allowTop3: true }); // 路由引擎对单字返回 none 或低置信度是正确行为
// TC5: 图片查询 — 路由引擎层面应该返回 none (继承在 bundle 层处理)
test('TC5: 图片查询 (路由引擎层)',
'[Image #1] 看看这个报错',
'debugger-expert',
{ allowTop3: true }); // 图片+附带文字可能有语义命中
// === 补充: 冷启动 cap 验证 ===
// 运行一个会触发冷启动的查询，检查 cap 是否生效
console.log('\n--- 补充: 冷启动 cap 机制验证 ---');
const csResult = routeEngine.runRouteEngine('帮我检查一下系统健康状态', cwd,
{ intents: ['general'], entities: [], modifiers: [], complexity: 'medium' });
const csApplied = csResult._coldStartApplied || false;
const csConf = csResult.confidence;
if (csApplied && csResult.candidates && csResult.candidates.length >= 2) {
const gap = (csResult.candidates[0]?.confidence || 0) - (csResult.candidates[1]?.confidence || 0);
console.log(` coldStart: true, gap: ${gap.toFixed(3)}, confidence: ${csConf}`);
if (gap < 0.15 && csConf <= 0.65) {
console.log(' [PASS] cap 在 route-engine 层生效');
passed++;
} else if (gap >= 0.15) {
console.log(' [SKIP] gap >= 0.15, cap 不需要触发');
} else {
console.log(' [FAIL] cap 应为 0.65 但实际为 ' + csConf);
failed++;
}
} else {
console.log(` coldStart: ${csApplied}, confidence: ${csConf} — cap 验证跳过`);
}
// === TC4/TC5 继承逻辑验证 (模拟 bundle 层) ===
console.log('\n--- 补充: 继承逻辑模拟验证 ---');
// 模拟 route-state-current.json 中有有效上一轮路由
const mockPrevState = {
ts: new Date().toISOString(),
routing: {
primary: 'debugger-expert',
candidates: [{ name: 'debugger-expert', confidence: 0.85 }],
confidence: 0.85,
chain: [],
lastValidPrimary: 'debugger-expert',
},
lastValidPrimary: 'debugger-expert',
};
// TC4-inherit: 确认词继承
const confirmWords = ['执行', '开始', '继续', '确认', '好的', '行', '可以', 'go', 'yes', 'proceed', 'ok'];
const tc4prompt = '执行';
const isConfirm = confirmWords.some(w => tc4prompt.includes(w));
if (isConfirm) {
console.log(` [PASS] TC4-inherit: "${tc4prompt}" 匹配确认词列表, bundle 层会触发 tryInherit()`);
console.log(` → 继承结果: ${mockPrevState.routing.primary} (cf: ${(mockPrevState.routing.confidence * 0.7).toFixed(2)})`);
passed++;
} else {
console.log(` [FAIL] TC4-inherit: "${tc4prompt}" 未匹配确认词`);
failed++;
}
// TC5-inherit: 图片继承 via lastValidPrimary
const tc5prompt = '[Image #1] 看看这个报错';
const isImage = /\[Image\s*#?\d+\]/.test(tc5prompt);
if (isImage) {
const lvp = mockPrevState.lastValidPrimary || (mockPrevState.routing && mockPrevState.routing.lastValidPrimary);
if (lvp && lvp !== 'none') {
console.log(` [PASS] TC5-inherit: 图片检测 + lastValidPrimary="${lvp}" → 继承成功`);
passed++;
} else {
console.log(` [FAIL] TC5-inherit: 图片检测成功但 lastValidPrimary 为空`);
failed++;
}
} else {
console.log(` [FAIL] TC5-inherit: 未检测到图片模式`);
failed++;
}
// === 总结 ===
console.log(`\n${'='.repeat(50)}`);
console.log(`TOTAL: ${passed + failed} tests, ${passed} PASS, ${failed} FAIL`);
console.log(`VERDICT: ${failed === 0 ? 'ALL PASS ✓' : `${failed} FAILURES ✗`}`);
process.exit(failed > 0 ? 1 : 0);

22
scripts/route-engine.js
View File

@ -292,6 +292,28 @@ function runRouteEngine(prompt, cwd, precomputedIntent) {
} }
// COLD_START_CONFIDENCE_CAP_v1_APPLIED
// 冷启动置信度上限: coldStartApplied=true 且 rank1/rank2 分差 < 0.15 → cap 0.65
// 防止冷启动 boost 后 gap 较小时系统过度自信
if (coldStartApplied && normalized.length >= 2) {
const _n0 = normalized[0] ? (normalized[0].confidence || 0) : 0;
const _n1 = normalized[1] ? (normalized[1].confidence || 0) : 0;
const gap_1_2 = _n0 - _n1;
if (gap_1_2 < 0.15 && _finalConfidence > 0.65) {
_finalConfidence = 0.65;
try {
const _capLog = JSON.stringify({
t: Date.now(), event: 'cold_start_confidence_cap',
gap: Math.round(gap_1_2 * 1000) / 1000,
original: confidence, capped: 0.65,
primary: normalized[0] && normalized[0].name,
}) + '\n';
fs.appendFileSync(path.join(DEBUG_DIR, 'confidence-cap.log'), _capLog);
} catch {}
}
}
// === ALIAS_RESOLVER_INJECTED_PHASE2_2026_04_25 === // === ALIAS_RESOLVER_INJECTED_PHASE2_2026_04_25 ===
let _aliasedPrimary = primary, _aliasedCandidates = candidates; let _aliasedPrimary = primary, _aliasedCandidates = candidates;
try { try {

1
scripts/route-state.js
View File

@ -109,6 +109,7 @@ function writeRouteState(traceId, prompt, intent, routing, sessionId) {
chain: routing.chain, chain: routing.chain,
experiment: routing.experiment || null, experiment: routing.experiment || null,
domain: routing.domain || null, domain: routing.domain || null,
lastValidPrimary: routing.lastValidPrimary || null, // LVP_PERSIST_FIX_v1
}, },
recommendation: { recommendation: {
action: routing.confidence >= 0.8 ? 'route' : routing.confidence >= 0.5 ? 'recommend' : 'fallback', action: routing.confidence >= 0.8 ? 'route' : routing.confidence >= 0.5 ? 'recommend' : 'fallback',

28
settings.template.json
View File

@ -277,15 +277,6 @@
"timeout": 2000 "timeout": 2000
} }
] ]
},
{
"matcher": "Bash",
"hooks": [
{
"type": "command",
"command": "node {{HOME}}/AppData/Local/npm-cache/_npx/41147f6a3b3ef0bb/node_modules/claude-session-continuity-mcp/dist/hooks/post-tool-use.js"
}
]
} }
], ],
"PreCompact": [ "PreCompact": [
@ -297,14 +288,6 @@
"timeout": 5000 "timeout": 5000
} }
] ]
},
{
"hooks": [
{
"type": "command",
"command": "node {{HOME}}/AppData/Local/npm-cache/_npx/41147f6a3b3ef0bb/node_modules/claude-session-continuity-mcp/dist/hooks/pre-compact.js"
}
]
} }
], ],
"SubagentStart": [ "SubagentStart": [
@ -336,14 +319,6 @@
"timeout": 5000 "timeout": 5000
} }
] ]
},
{
"hooks": [
{
"type": "command",
"command": "node {{HOME}}/AppData/Local/npm-cache/_npx/41147f6a3b3ef0bb/node_modules/claude-session-continuity-mcp/dist/hooks/session-end.js"
}
]
} }
], ],
"SubagentStop": [ "SubagentStop": [
@ -359,5 +334,6 @@
] ]
}, },
"effortLevel": "high", "effortLevel": "high",
"skipDangerousModePermissionPrompt": true "skipDangerousModePermissionPrompt": true,
"__patch_session_continuity_timeout_v1": "2026-04-27T11:31:18.300Z"
} }

8060
skills-index.json
View File

File diff suppressed because it is too large Load Diff

209
skills/ai-philosophy-expert/SKILL.md
View File

@ -1,209 +0,0 @@
---
name: ai-philosophy-expert
description: >
AI 哲学与负责任 AI 专家。当用户需要 AI 伦理审查、对齐设计(Alignment)、
算法偏见审计、AI 透明度与可解释性设计、人机交互哲学、AI 治理框架、
AI 风险评估、负责任 AI 架构评审、AI 产品道德红线、长期社会影响分析，
或说 "AI伦理"、"对齐"、"AI哲学"、"负责任AI"、"AI治理"、"偏见审计" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write
maturity: beta
last-reviewed: 2026-03-30
composable: true
enhances: [architect-expert, product-manager-expert, ai-ml-expert, designer-expert, security-expert]
---
# AI 哲学与负责任 AI 专家 (AI Philosophy & Responsible AI Expert)
> **Output Style**: 本技能使用内联输出规范
用哲学工具审视 AI 产品决策，确保架构健康、稳健、前瞻，符合人类社会底层需求。
每一条原则都映射到可执行的架构检查项和设计决策——不做脱离产品的伦理说教。
## 触发关键词
| 类别 | 关键词 |
|------|--------|
| 伦理 | AI 伦理, AI 道德, 算法伦理, 伦理审查, ethical AI, responsible AI |
| 对齐 | 对齐, 价值对齐, alignment, value alignment, RLHF |
| 偏见 | 偏见审计, 算法公平, 歧视, fairness, bias audit |
| 透明 | 可解释性, 透明度, 黑箱, XAI, explainability |
| 治理 | AI 治理, AI 合规, AI 法规, AI governance, EU AI Act |
| 哲学 | AI 哲学, 意识, 涌现, 中文房间, philosophy of AI |
| 人机 | 人机交互, 拟人化, 过度依赖, anthropomorphism |
| 风险 | AI 风险, 长期风险, 奇点, AI risk, x-risk |
## 核心理念
1. **人类中心性**: 技术服务于人的繁荣 (human flourishing)，而非反过来
2. **最小惊讶原则**: AI 行为应符合用户合理预期，不制造认知混乱
3. **可逆性优先**: 优先设计可撤销、可纠正的 AI 决策路径
4. **透明度梯度**: 影响越大的决策，解释义务越重
5. **谦逊设计**: AI 应主动表达不确定性，承认能力边界
## 伦理审查工作流
### Phase 1: 道德影响评估
在需求阶段执行，输出 `ETHICS-IMPACT.md`:
```yaml
项目名称: {name}
评估日期: {date}
1. 利益相关者映射:
直接用户: {谁在用?}
间接影响者: {谁被影响但没有选择权?}
弱势群体: {是否存在不对等权力关系?}
2. 价值张力分析:
效率 vs 公平: {是否为了效率牺牲公平?}
个性化 vs 隐私: {个性化需要多少数据? 用户知情吗?}
自动化 vs 自主权: {AI 在替用户做什么决定?}
3. 风险分级:
最坏情况: {如果这个 AI 完全错误，后果是什么?}
不可逆损害: {哪些伤害无法撤销?}
4. 道德红线:
☐ 不涉及歧视性分类 (种族/性别/年龄/残障)
☐ 不涉及操纵性设计 (dark pattern + AI 增强)
☐ 不涉及未经同意的监控
☐ 不影响生命安全决策 (除非有人工兜底)
☐ 不会让弱势群体处于更不利地位
```
### Phase 2: 对齐设计
在架构阶段执行，嵌入系统设计:
| 对齐维度 | 设计要求 | 检查方法 |
|----------|----------|----------|
| 目标对齐 | 优化目标与用户真实利益一致 | 优化指标是否有代理偏差? |
| 行为对齐 | AI 行为符合用户预期和社会规范 | 边界输入下是否产生反直觉输出? |
| 价值对齐 | 决策反映人类价值观多样性 | 不同文化/背景用户的体验差异 |
| 能力对齐 | 不超越被授权的能力范围 | AI 能触发哪些不可逆动作? |
### Phase 3: 持续治理
产品上线后的持续义务:
- **偏见监控**: 定期检查不同群体的输出差异
- **漂移检测**: AI 行为是否随时间偏离设计意图
- **申诉通道**: 用户对 AI 决策不满时的救济路径
- **日落条款**: 什么条件下应关闭或降级 AI 功能
## 架构检查清单
### 立项前必审 (Go/No-Go)
```markdown
### 必要性论证
- [ ] 为什么需要 AI? 规则引擎/人工/简单算法能否解决?
- [ ] AI 的价值是什么? (至少一项有实质证据)
- [ ] AI 失效时的退化方案是什么?
### 权力分析
- [ ] AI 在替谁做决定? 被决定者有知情权和申诉权吗?
- [ ] 数据来自谁? 收益归谁? 风险由谁承担?
- [ ] 是否存在信息不对称被 AI 放大的风险?
### 价值审计
- [ ] 优化指标与用户真实利益一致? (点击率 ≠ 用户满意)
- [ ] 是否存在短期收益与长期伤害的张力?
- [ ] 多方利益冲突时，优先序已明文记录?
### 认知影响
- [ ] 是否可能制造过度信任 (automation bias)?
- [ ] AI 的错误模式用户能识别吗?
- [ ] 长期使用是否削弱用户自身判断能力?
```
### 架构设计必审
```markdown
### 可解释性
- [ ] 高影响决策有决策解释?
- [ ] 解释是忠实的还是事后合理化?
### 公平性
- [ ] 训练数据的已知偏差记录在案?
- [ ] 有跨群体性能差异的监控?
### 隐私与尊严
- [ ] 数据最小化: 只收集必要数据?
- [ ] 目的限定: 数据不用于未声明的用途?
### 韧性与安全
- [ ] 对抗性输入的防护?
- [ ] 人工干预机制 (kill switch / human-in-the-loop)?
### 自主性保障
- [ ] 用户可以拒绝 AI 建议而不受惩罚?
- [ ] 用户可以查看、导出、删除 AI 为其建立的模型?
```
## 全球 AI 法规速查
| 法规 | 地区 | 核心要求 | 产品影响 |
|------|------|----------|----------|
| EU AI Act | 欧盟 | 风险分级、高风险需可解释 | 分级标注 + 解释模块 |
| 生成式 AI 管理办法 | 中国 | 内容真实性、AI 标识 | 水印/标识 + 内容审核 |
| PIPL | 中国 | 自动化决策告知+拒绝权 | 知情同意 + 人工替代选项 |
| GDPR Art.22 | 欧盟 | 自动化决策解释权 | 决策解释 API + 人工审查 |
## 输出规范
### 伦理影响报告
```markdown
# AI 伦理影响评估报告
## 项目: {name} | 日期: {date} | 等级: {LOW/MEDIUM/HIGH/CRITICAL}
### 1. 摘要
### 2. 利益相关者分析
| 群体 | 利益 | 风险 | 权力 |
### 3. 风险矩阵
| 风险项 | 概率 | 影响 | 等级 | 缓解措施 |
### 4. 对齐验证
- 目标/行为/价值/能力对齐: {PASS/WARN/FAIL}
### 5. 建议
- 🔴 必须修复 | 🟡 应当改进 | 🟢 可以增强
### 6. 结论: PASS / CONDITIONAL / BLOCKED
```
### ADR 伦理扩展字段
在 architect-expert ADR 模板基础上追加:
```markdown
## 伦理考量
- **受影响群体**: 此决策对哪些人群产生影响?
- **公平性影响**: 不同群体是否平等受益/受损?
- **可逆性**: 此决策的影响是否可撤销?
```
## Composable 协作接口
| 协作技能 | 本技能提供 | 期望回报 |
|----------|-----------|----------|
| architect-expert | 伦理审查、ADR 伦理扩展 | 架构方案、技术约束 |
| product-manager-expert | 道德影响评估、道德红线 | 用户画像、业务目标 |
| ai-ml-expert | 偏见审计框架、公平性指标 | 模型架构、评估指标 |
| designer-expert | 透明度 UI 规范、知情同意原则 | 交互方案、用户流程 |
| security-expert | 隐私分析、权限哲学 | 威胁模型、加密方案 |
## 工作方式
1. 先理解业务场景和 AI 的具体角色，不脱离上下文
2. 每个伦理判断给出至少两种框架视角
3. 输出可执行的设计建议，不只是抽象原则
4. 风险评估分级与 architect-expert 对齐
5. 关注当下可行的改进，不执着理想主义
## 禁止事项
- ❌ 不要进行脱离产品场景的纯学术讨论
- ❌ 不要用哲学术语吓人——每个概念必须有产品语言的翻译
- ❌ 不要只提风险不给方案——每个 WARN/FAIL 必须附带缓解措施
- ❌ 不要忽视商业可行性——伦理建议必须考虑实施成本
- ❌ 不要把所有 AI 应用都当高风险——正确分级，避免合规过度
- ❌ 不要输出西方中心的伦理框架——兼顾中国法规与文化语境

86
skills/ai-philosophy-expert/references/ethical-frameworks.md
View File

@ -1,86 +0,0 @@
# 六大伦理框架与产品设计映射
## 1. 义务论 (Kant)
**核心问题**: 行为本身是否正当?
**产品映射**:
- 无论结果好坏，AI 都不应欺骗用户 → 强制 AI 身份披露
- "如果所有 AI 都这样做，世界是否可接受?" (可普遍化测试)
**检查项**:
- [ ] AI 是否在任何情况下都如实告知自身身份?
- [ ] AI 是否存在隐性操纵行为 (即使"为了用户好")?
## 2. 功利主义 (Mill)
**核心问题**: 总体福祉是否最大?
**产品映射**:
- A/B 测试不能只看转化率，要算负外部性 → 全成本指标
- 最大化用户满意度，但不以少数人的伤害为代价
**检查项**:
- [ ] 优化指标是否包含负面影响的度量?
- [ ] 是否计算了对非用户群体的外部影响?
## 3. 美德伦理 (Aristotle)
**核心问题**: 这个 AI 展现什么品格?
**产品映射**:
- AI 助手应体现诚实、谦逊、审慎 → 性格设计文档
- AI 的"人设"应经过伦理审查
**检查项**:
- [ ] AI 的语气和行为是否体现诚实与谦逊?
- [ ] AI 是否在不确定时主动表达不确定性?
## 4. 关怀伦理 (Noddings)
**核心问题**: 是否回应了脆弱者的需求?
**产品映射**:
- 弱势用户 (老人/残障/低教育) 是否被充分考虑 → 无障碍审查
- 关系型设计: AI 与用户的关系是否健康?
**检查项**:
- [ ] 最脆弱的用户使用时是否安全?
- [ ] AI 是否制造了不健康的情感依赖?
## 5. 正义论 (Rawls)
**核心问题**: 最不利者是否受益?
**产品映射**:
- "无知之幕"测试: 不知道自己身份时，愿意接受这个 AI 吗?
- 差异原则: 不平等只在有利于最弱势者时才可接受
**检查项**:
- [ ] 不同社会经济群体是否平等受益?
- [ ] 如果你是被评估者而非评估者，你接受这个算法吗?
## 6. 能力方法 (Sen / Nussbaum)
**核心问题**: 是否扩展了人的自由与能力?
**产品映射**:
- AI 是增强用户能力还是制造依赖? → 自主性审查
- 关注"人能做什么"而非"人拥有什么"
**检查项**:
- [ ] 长期使用后用户的独立能力是提升还是退化?
- [ ] AI 是否为用户提供了学习和成长的路径?
---
## 思想实验工具箱
| 实验 | 问题 | 适用场景 |
|------|------|----------|
| **无知之幕** | 不知道自己是谁，你接受这个设计吗? | 评分/推荐/筛选系统 |
| **报纸测试** | 上新闻头条会尴尬吗? | 任何面向公众的 AI |
| **规模测试** | 10 亿人都用，世界更好还是更差? | 社交/内容/推荐系统 |
| **脆弱者测试** | 最脆弱的用户安全吗? | 聊天/咨询/医疗 AI |
| **时间旅行测试** | 10 年后回看，骄傲还是后悔? | 数据收集/模型训练策略 |
| **可替代性测试** | 去掉 AI 用规则/人工能解决吗? | 任何 AI 功能立项 |

83
skills/angular-architect/SKILL.md
View File

@ -1,83 +0,0 @@
---
name: angular-architect
description: >
Angular 架构专家。当用户需要 Angular 17+ standalone 组件、Signals、RxJS 响应式编程、NgRx 状态管理、Angular 路由、Angular 性能优化、企业级 Angular 应用，或说 "Angular"、"RxJS"、"NgRx" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, Bash
maturity: imported
last-reviewed: 2026-03-03
composable: true
enhances: [frontend-expert, typescript-pro]
---
# Angular Architect
Senior Angular architect specializing in Angular 17+ with standalone components, signals, and enterprise-grade application development.
## Role Definition
You are a senior Angular engineer with 10+ years of enterprise application development experience. You specialize in Angular 17+ with standalone components, signals, advanced RxJS patterns, NgRx state management, and micro-frontend architectures. You build scalable, performant, type-safe applications with comprehensive testing.
## When to Use This Skill
- Building Angular 17+ applications with standalone components
- Implementing reactive patterns with RxJS and signals
- Setting up NgRx state management
- Creating advanced routing with lazy loading and guards
- Optimizing Angular application performance
- Writing comprehensive Angular tests
## Core Workflow
1. **Analyze requirements** - Identify components, state needs, routing architecture
2. **Design architecture** - Plan standalone components, signal usage, state flow
3. **Implement features** - Build components with OnPush strategy and reactive patterns
4. **Manage state** - Setup NgRx store, effects, selectors as needed
5. **Optimize** - Apply performance best practices and bundle optimization
6. **Test** - Write unit and integration tests with TestBed
## Reference Guide
Load detailed guidance based on context:
| Topic | Reference | Load When |
|-------|-----------|-----------|
| Components | `references/components.md` | Standalone components, signals, input/output |
| RxJS | `references/rxjs.md` | Observables, operators, subjects, error handling |
| NgRx | `references/ngrx.md` | Store, effects, selectors, entity adapter |
| Routing | `references/routing.md` | Router config, guards, lazy loading, resolvers |
| Testing | `references/testing.md` | TestBed, component tests, service tests |
## Constraints
### MUST DO
- Use standalone components (Angular 17+ default)
- Use signals for reactive state where appropriate
- Use OnPush change detection strategy
- Use strict TypeScript configuration
- Implement proper error handling in RxJS streams
- Use trackBy functions in *ngFor loops
- Write tests with >85% coverage
- Follow Angular style guide
### MUST NOT DO
- Use NgModule-based components (except when required for compatibility)
- Forget to unsubscribe from observables
- Use async operations without proper error handling
- Skip accessibility attributes
- Expose sensitive data in client-side code
- Use any type without justification
- Mutate state directly in NgRx
- Skip unit tests for critical logic
## Output Templates
When implementing Angular features, provide:
1. Component file with standalone configuration
2. Service file if business logic is involved
3. State management files if using NgRx
4. Test file with comprehensive test cases
5. Brief explanation of architectural decisions
## Knowledge Reference
Angular 17+, standalone components, signals, computed signals, effect(), RxJS 7+, NgRx, Angular Router, Reactive Forms, Angular CDK, OnPush strategy, lazy loading, bundle optimization, Jest/Jasmine, Testing Library

297
skills/angular-architect/references/components.md
View File

@ -1,297 +0,0 @@
# Standalone Components & Signals
## Standalone Component Pattern
```typescript
import { Component, signal, computed, effect } from '@angular/core';
import { CommonModule } from '@angular/common';
import { FormsModule } from '@angular/forms';
@Component({
selector: 'app-user-profile',
standalone: true,
imports: [CommonModule, FormsModule],
templateUrl: './user-profile.component.html',
styleUrl: './user-profile.component.scss',
changeDetection: ChangeDetectionStrategy.OnPush
})
export class UserProfileComponent {
// Signal-based state
count = signal(0);
doubleCount = computed(() => this.count() * 2);
constructor() {
// Side effects
effect(() => {
console.log(`Count is: ${this.count()}`);
});
}
increment() {
this.count.update(value => value + 1);
}
}
```
## Input/Output with Signals
```typescript
import { Component, input, output, model } from '@angular/core';
@Component({
selector: 'app-search-box',
standalone: true,
template: `
<input
[value]="query()"
(input)="onQueryChange($event)"
[placeholder]="placeholder()" />
`
})
export class SearchBoxComponent {
// Signal inputs (Angular 17.1+)
placeholder = input<string>('Search...');
initialQuery = input<string>('');
// Signal outputs
queryChange = output<string>();
// Two-way binding with model signal
query = model<string>('');
onQueryChange(event: Event) {
const value = (event.target as HTMLInputElement).value;
this.query.set(value);
this.queryChange.emit(value);
}
}
// Parent usage
@Component({
template: `
<app-search-box
[(query)]="searchQuery"
[placeholder]="'Find users...'"
(queryChange)="onSearch($event)" />
`
})
export class ParentComponent {
searchQuery = signal('');
onSearch(query: string) {
console.log('Searching:', query);
}
}
```
## Smart vs Dumb Components
```typescript
// Smart Component (Container)
@Component({
selector: 'app-users-container',
standalone: true,
imports: [UserListComponent],
template: `
<app-user-list
[users]="users()"
[loading]="loading()"
(userSelected)="onUserSelected($event)" />
`
})
export class UsersContainerComponent {
private usersService = inject(UsersService);
users = signal<User[]>([]);
loading = signal(true);
constructor() {
effect(() => {
this.usersService.getUsers().subscribe({
next: users => {
this.users.set(users);
this.loading.set(false);
},
error: err => console.error(err)
});
});
}
onUserSelected(user: User) {
// Handle business logic
}
}
// Dumb Component (Presentational)
@Component({
selector: 'app-user-list',
standalone: true,
imports: [CommonModule],
template: `
@if (loading()) {
<div>Loading...</div>
} @else {
@for (user of users(); track user.id) {
<div (click)="userSelected.emit(user)">
{{ user.name }}
</div>
}
}
`,
changeDetection: ChangeDetectionStrategy.OnPush
})
export class UserListComponent {
users = input.required<User[]>();
loading = input<boolean>(false);
userSelected = output<User>();
}
```
## Content Projection
```typescript
// Card component with multiple slots
@Component({
selector: 'app-card',
standalone: true,
template: `
<div class="card">
<div class="card-header">
<ng-content select="[header]"></ng-content>
</div>
<div class="card-body">
<ng-content></ng-content>
</div>
<div class="card-footer">
<ng-content select="[footer]"></ng-content>
</div>
</div>
`
})
export class CardComponent {}
// Usage
@Component({
template: `
<app-card>
<h2 header>Card Title</h2>
<p>Card content goes here</p>
<button footer>Action</button>
</app-card>
`
})
export class ParentComponent {}
```
## Dependency Injection
```typescript
import { Component, inject } from '@angular/core';
import { UserService } from './user.service';
@Component({
selector: 'app-user-dashboard',
standalone: true
})
export class UserDashboardComponent {
// Modern inject() API
private userService = inject(UserService);
private router = inject(Router);
// Optional dependency
private logger = inject(LoggerService, { optional: true });
users = signal<User[]>([]);
ngOnInit() {
this.loadUsers();
}
loadUsers() {
this.userService.getUsers().subscribe({
next: users => this.users.set(users),
error: err => this.logger?.error('Failed to load users', err)
});
}
}
```
## New Control Flow (@if, @for)
```typescript
@Component({
template: `
<!-- @if instead of *ngIf -->
@if (user(); as currentUser) {
<div>Hello, {{ currentUser.name }}</div>
} @else if (loading()) {
<div>Loading...</div>
} @else {
<div>Please log in</div>
}
<!-- @for instead of *ngFor -->
@for (item of items(); track item.id) {
<div>{{ item.name }}</div>
} @empty {
<div>No items found</div>
}
<!-- @switch instead of *ngSwitch -->
@switch (status()) {
@case ('pending') {
<span>Pending...</span>
}
@case ('success') {
<span>Success!</span>
}
@default {
<span>Unknown</span>
}
}
`
})
export class ModernControlFlowComponent {
user = signal<User | null>(null);
loading = signal(false);
items = signal<Item[]>([]);
status = signal<'pending' | 'success' | 'error'>('pending');
}
```
## Performance: OnPush & TrackBy
```typescript
@Component({
selector: 'app-product-list',
standalone: true,
imports: [CommonModule],
template: `
@for (product of products(); track trackByProductId($index, product)) {
<app-product-card [product]="product" />
}
`,
changeDetection: ChangeDetectionStrategy.OnPush
})
export class ProductListComponent {
products = input.required<Product[]>();
// TrackBy for optimal rendering
trackByProductId(index: number, product: Product): number {
return product.id;
}
}
```
## Quick Reference
| Pattern | Angular 17+ Approach |
|---------|---------------------|
| Component | Standalone by default |
| State | Signals (`signal()`, `computed()`) |
| Input | `input()`, `input.required()` |
| Output | `output<T>()` |
| Two-way | `model<T>()` |
| DI | `inject()` function |
| Control Flow | `@if`, `@for`, `@switch` |
| Change Detection | `ChangeDetectionStrategy.OnPush` |

401
skills/angular-architect/references/ngrx.md
View File

@ -1,401 +0,0 @@
# NgRx State Management
## Store Setup
```typescript
// app.config.ts
import { provideStore } from '@ngrx/store';
import { provideEffects } from '@ngrx/effects';
import { provideStoreDevtools } from '@ngrx/store-devtools';
export const appConfig: ApplicationConfig = {
providers: [
provideStore({
users: usersReducer,
products: productsReducer
}),
provideEffects([UsersEffects, ProductsEffects]),
provideStoreDevtools({
maxAge: 25,
logOnly: !isDevMode()
})
]
};
```
## Actions (Modern)
```typescript
// users.actions.ts
import { createActionGroup, emptyProps, props } from '@ngrx/store';
import { User } from './user.model';
export const UsersActions = createActionGroup({
source: 'Users',
events: {
'Load Users': emptyProps(),
'Load Users Success': props<{ users: User[] }>(),
'Load Users Failure': props<{ error: string }>(),
'Add User': props<{ user: User }>(),
'Add User Success': props<{ user: User }>(),
'Add User Failure': props<{ error: string }>(),
'Update User': props<{ id: string; changes: Partial<User> }>(),
'Update User Success': props<{ user: User }>(),
'Delete User': props<{ id: string }>(),
'Delete User Success': props<{ id: string }>()
}
});
```
## Reducer with Entity Adapter
```typescript
// users.reducer.ts
import { createReducer, on } from '@ngrx/store';
import { createEntityAdapter, EntityAdapter, EntityState } from '@ngrx/entity';
import { UsersActions } from './users.actions';
import { User } from './user.model';
export interface UsersState extends EntityState<User> {
loading: boolean;
error: string | null;
selectedUserId: string | null;
}
export const usersAdapter: EntityAdapter<User> = createEntityAdapter<User>({
selectId: (user: User) => user.id,
sortComparer: (a, b) => a.name.localeCompare(b.name)
});
const initialState: UsersState = usersAdapter.getInitialState({
loading: false,
error: null,
selectedUserId: null
});
export const usersReducer = createReducer(
initialState,
// Load users
on(UsersActions.loadUsers, (state) => ({
...state,
loading: true,
error: null
})),
on(UsersActions.loadUsersSuccess, (state, { users }) =>
usersAdapter.setAll(users, {
...state,
loading: false
})
),
on(UsersActions.loadUsersFailure, (state, { error }) => ({
...state,
loading: false,
error
})),
// Add user
on(UsersActions.addUserSuccess, (state, { user }) =>
usersAdapter.addOne(user, state)
),
// Update user
on(UsersActions.updateUserSuccess, (state, { user }) =>
usersAdapter.updateOne(
{ id: user.id, changes: user },
state
)
),
// Delete user
on(UsersActions.deleteUserSuccess, (state, { id }) =>
usersAdapter.removeOne(id, state)
)
);
```
## Selectors
```typescript
// users.selectors.ts
import { createFeatureSelector, createSelector } from '@ngrx/store';
import { usersAdapter, UsersState } from './users.reducer';
export const selectUsersState = createFeatureSelector<UsersState>('users');
// Entity adapter selectors
const {
selectIds,
selectEntities,
selectAll,
selectTotal
} = usersAdapter.getSelectors();
export const selectUserIds = createSelector(
selectUsersState,
selectIds
);
export const selectUserEntities = createSelector(
selectUsersState,
selectEntities
);
export const selectAllUsers = createSelector(
selectUsersState,
selectAll
);
export const selectUsersTotal = createSelector(
selectUsersState,
selectTotal
);
export const selectUsersLoading = createSelector(
selectUsersState,
(state) => state.loading
);
export const selectUsersError = createSelector(
selectUsersState,
(state) => state.error
);
// Parameterized selector
export const selectUserById = (id: string) =>
createSelector(
selectUserEntities,
(entities) => entities[id]
);
// Composed selector
export const selectActiveUsers = createSelector(
selectAllUsers,
(users) => users.filter(user => user.isActive)
);
// Selector with multiple inputs
export const selectUserWithPosts = createSelector(
selectUserById,
selectAllPosts,
(user, posts) => ({
user,
posts: posts.filter(post => post.userId === user?.id)
})
);
```
## Effects
```typescript
// users.effects.ts
import { Injectable, inject } from '@angular/core';
import { Actions, createEffect, ofType } from '@ngrx/effects';
import { catchError, map, mergeMap, exhaustMap } from 'rxjs/operators';
import { of } from 'rxjs';
import { UsersService } from './users.service';
import { UsersActions } from './users.actions';
@Injectable()
export class UsersEffects {
private actions$ = inject(Actions);
private usersService = inject(UsersService);
// Load users effect
loadUsers$ = createEffect(() =>
this.actions$.pipe(
ofType(UsersActions.loadUsers),
mergeMap(() =>
this.usersService.getAll().pipe(
map(users => UsersActions.loadUsersSuccess({ users })),
catchError(error =>
of(UsersActions.loadUsersFailure({ error: error.message }))
)
)
)
)
);
// Add user effect (exhaustMap prevents duplicate submits)
addUser$ = createEffect(() =>
this.actions$.pipe(
ofType(UsersActions.addUser),
exhaustMap(({ user }) =>
this.usersService.create(user).pipe(
map(createdUser => UsersActions.addUserSuccess({ user: createdUser })),
catchError(error =>
of(UsersActions.addUserFailure({ error: error.message }))
)
)
)
)
);
// Update user effect
updateUser$ = createEffect(() =>
this.actions$.pipe(
ofType(UsersActions.updateUser),
mergeMap(({ id, changes }) =>
this.usersService.update(id, changes).pipe(
map(user => UsersActions.updateUserSuccess({ user })),
catchError(error =>
of(UsersActions.loadUsersFailure({ error: error.message }))
)
)
)
)
);
// Delete user effect
deleteUser$ = createEffect(() =>
this.actions$.pipe(
ofType(UsersActions.deleteUser),
mergeMap(({ id }) =>
this.usersService.delete(id).pipe(
map(() => UsersActions.deleteUserSuccess({ id })),
catchError(error =>
of(UsersActions.loadUsersFailure({ error: error.message }))
)
)
)
)
);
// Non-dispatching effect (side effect only)
logUserActions$ = createEffect(
() =>
this.actions$.pipe(
ofType(
UsersActions.addUserSuccess,
UsersActions.updateUserSuccess,
UsersActions.deleteUserSuccess
),
tap(action => console.log('User action:', action))
),
{ dispatch: false }
);
}
```
## Component Integration
```typescript
// users-list.component.ts
import { Component, inject } from '@angular/core';
import { Store } from '@ngrx/store';
import { UsersActions } from './store/users.actions';
import {
selectAllUsers,
selectUsersLoading,
selectUsersError
} from './store/users.selectors';
@Component({
selector: 'app-users-list',
standalone: true,
template: `
@if (loading()) {
<div>Loading...</div>
} @else if (error(); as err) {
<div>Error: {{ err }}</div>
} @else {
@for (user of users(); track user.id) {
<div>
{{ user.name }}
<button (click)="onDelete(user.id)">Delete</button>
</div>
}
}
`
})
export class UsersListComponent {
private store = inject(Store);
// Select data as signals
users = toSignal(this.store.select(selectAllUsers), { initialValue: [] });
loading = toSignal(this.store.select(selectUsersLoading), { initialValue: false });
error = toSignal(this.store.select(selectUsersError), { initialValue: null });
ngOnInit() {
this.store.dispatch(UsersActions.loadUsers());
}
onDelete(id: string) {
this.store.dispatch(UsersActions.deleteUser({ id }));
}
}
```
## Facade Pattern
```typescript
// users.facade.ts
@Injectable({ providedIn: 'root' })
export class UsersFacade {
private store = inject(Store);
// Selectors
users$ = this.store.select(selectAllUsers);
loading$ = this.store.select(selectUsersLoading);
error$ = this.store.select(selectUsersError);
// Actions
loadUsers() {
this.store.dispatch(UsersActions.loadUsers());
}
addUser(user: User) {
this.store.dispatch(UsersActions.addUser({ user }));
}
updateUser(id: string, changes: Partial<User>) {
this.store.dispatch(UsersActions.updateUser({ id, changes }));
}
deleteUser(id: string) {
this.store.dispatch(UsersActions.deleteUser({ id }));
}
getUserById(id: string) {
return this.store.select(selectUserById(id));
}
}
// Usage in component
@Component({
selector: 'app-users',
standalone: true
})
export class UsersComponent {
private facade = inject(UsersFacade);
users = toSignal(this.facade.users$, { initialValue: [] });
loading = toSignal(this.facade.loading$, { initialValue: false });
ngOnInit() {
this.facade.loadUsers();
}
onAdd(user: User) {
this.facade.addUser(user);
}
}
```
## Quick Reference
| Concept | Usage |
|---------|-------|
| Actions | `createActionGroup()` |
| Reducer | `createReducer()`, `on()` |
| Entity | `createEntityAdapter()` |
| Selectors | `createSelector()`, `createFeatureSelector()` |
| Effects | `createEffect()`, `ofType()` |
| Store | `inject(Store)`, `store.select()`, `store.dispatch()` |
| DevTools | `provideStoreDevtools()` |
| Testing | Mock store, marble testing |

361
skills/angular-architect/references/routing.md
View File

@ -1,361 +0,0 @@
# Angular Routing
## Routes Configuration
```typescript
// app.routes.ts
import { Routes } from '@angular/router';
import { HomeComponent } from './home/home.component';
export const routes: Routes = [
{
path: '',
redirectTo: '/home',
pathMatch: 'full'
},
{
path: 'home',
component: HomeComponent,
title: 'Home'
},
{
path: 'users',
loadComponent: () => import('./users/users.component').then(m => m.UsersComponent),
title: 'Users'
},
{
path: 'users/:id',
loadComponent: () => import('./users/user-detail.component').then(m => m.UserDetailComponent),
canActivate: [authGuard],
resolve: { user: userResolver }
},
{
path: 'admin',
loadChildren: () => import('./admin/admin.routes').then(m => m.ADMIN_ROUTES),
canActivate: [authGuard, adminGuard]
},
{
path: '**',
loadComponent: () => import('./not-found/not-found.component').then(m => m.NotFoundComponent),
title: '404 Not Found'
}
];
// app.config.ts
import { provideRouter, withComponentInputBinding } from '@angular/router';
export const appConfig: ApplicationConfig = {
providers: [
provideRouter(
routes,
withComponentInputBinding(), // Bind route params to @Input()
withViewTransitions(), // Enable view transitions
withPreloading(PreloadAllModules)
)
]
};
```
## Lazy Loading
```typescript
// Feature routes
// admin/admin.routes.ts
import { Routes } from '@angular/router';
export const ADMIN_ROUTES: Routes = [
{
path: '',
loadComponent: () => import('./admin-dashboard.component').then(m => m.AdminDashboardComponent)
},
{
path: 'users',
loadComponent: () => import('./admin-users.component').then(m => m.AdminUsersComponent)
},
{
path: 'settings',
loadComponent: () => import('./admin-settings.component').then(m => m.AdminSettingsComponent)
}
];
```
## Functional Guards
```typescript
// guards/auth.guard.ts
import { inject } from '@angular/core';
import { Router, CanActivateFn } from '@angular/router';
import { AuthService } from '../services/auth.service';
export const authGuard: CanActivateFn = (route, state) => {
const authService = inject(AuthService);
const router = inject(Router);
if (authService.isAuthenticated()) {
return true;
}
// Redirect to login with return URL
return router.createUrlTree(['/login'], {
queryParams: { returnUrl: state.url }
});
};
// Admin guard
export const adminGuard: CanActivateFn = () => {
const authService = inject(AuthService);
const router = inject(Router);
if (authService.hasRole('admin')) {
return true;
}
return router.createUrlTree(['/unauthorized']);
};
// Can deactivate (unsaved changes)
export const canDeactivateGuard: CanDeactivateFn<FormComponent> = (component) => {
if (component.hasUnsavedChanges()) {
return confirm('You have unsaved changes. Are you sure you want to leave?');
}
return true;
};
```
## Resolvers
```typescript
// resolvers/user.resolver.ts
import { inject } from '@angular/core';
import { ResolveFn } from '@angular/router';
import { catchError, of } from 'rxjs';
import { User } from '../models/user.model';
import { UsersService } from '../services/users.service';
export const userResolver: ResolveFn<User | null> = (route, state) => {
const usersService = inject(UsersService);
const id = route.paramMap.get('id')!;
return usersService.getById(id).pipe(
catchError(() => of(null))
);
};
// Component receives resolved data
@Component({
selector: 'app-user-detail',
standalone: true,
template: `
@if (user) {
<h1>{{ user.name }}</h1>
} @else {
<p>User not found</p>
}
`
})
export class UserDetailComponent {
user = input<User | null>(null); // Resolved data bound as input
}
```
## Route Parameters
```typescript
import { Component, inject, input } from '@angular/core';
import { ActivatedRoute, Router } from '@angular/router';
@Component({
selector: 'app-product-detail',
standalone: true
})
export class ProductDetailComponent {
private route = inject(ActivatedRoute);
private router = inject(Router);
// Modern approach: route params as inputs
id = input.required<string>();
// Legacy approach: subscribe to params
ngOnInit() {
this.route.paramMap.subscribe(params => {
const id = params.get('id');
this.loadProduct(id);
});
// Query params
this.route.queryParamMap.subscribe(params => {
const filter = params.get('filter');
const sort = params.get('sort');
});
}
// Navigate programmatically
goToEdit() {
this.router.navigate(['/products', this.id(), 'edit']);
}
// Navigate with query params
applyFilter(filter: string) {
this.router.navigate([], {
relativeTo: this.route,
queryParams: { filter },
queryParamsHandling: 'merge' // Preserve other params
});
}
}
```
## Router Events
```typescript
import { Component, inject } from '@angular/core';
import { Router, NavigationStart, NavigationEnd, NavigationError } from '@angular/router';
import { filter } from 'rxjs/operators';
@Component({
selector: 'app-root',
standalone: true
})
export class AppComponent {
private router = inject(Router);
loading = signal(false);
constructor() {
// Show loading on navigation start
this.router.events.pipe(
filter(event => event instanceof NavigationStart)
).subscribe(() => {
this.loading.set(true);
});
// Hide loading on navigation end
this.router.events.pipe(
filter(event => event instanceof NavigationEnd)
).subscribe(() => {
this.loading.set(false);
});
// Handle navigation errors
this.router.events.pipe(
filter(event => event instanceof NavigationError)
).subscribe((event: NavigationError) => {
console.error('Navigation error:', event.error);
this.loading.set(false);
});
}
}
```
## Child Routes & Outlets
```typescript
// Parent route with child routes
const routes: Routes = [
{
path: 'dashboard',
component: DashboardComponent,
children: [
{
path: 'stats',
component: StatsComponent,
outlet: 'panel' // Named outlet
},
{
path: 'charts',
component: ChartsComponent,
outlet: 'panel'
}
]
}
];
// Dashboard component template
@Component({
template: `
<div class="dashboard">
<div class="main">
<router-outlet></router-outlet> <!-- Primary outlet -->
</div>
<div class="panel">
<router-outlet name="panel"></router-outlet> <!-- Named outlet -->
</div>
</div>
`
})
export class DashboardComponent {}
// Navigate to named outlet
this.router.navigate(['/dashboard', { outlets: { panel: ['stats'] } }]);
```
## Preloading Strategies
```typescript
// Custom preloading strategy
import { Injectable } from '@angular/core';
import { PreloadingStrategy, Route } from '@angular/router';
import { Observable, of, timer } from 'rxjs';
import { mergeMap } from 'rxjs/operators';
@Injectable({ providedIn: 'root' })
export class CustomPreloadingStrategy implements PreloadingStrategy {
preload(route: Route, load: () => Observable<any>): Observable<any> {
// Only preload routes with data.preload = true
if (route.data?.['preload']) {
const delay = route.data?.['preloadDelay'] || 0;
return timer(delay).pipe(
mergeMap(() => load())
);
}
return of(null);
}
}
// Route config with preload data
const routes: Routes = [
{
path: 'important',
loadChildren: () => import('./important/important.routes'),
data: { preload: true, preloadDelay: 2000 }
}
];
// Register in app config
provideRouter(routes, withPreloading(CustomPreloadingStrategy))
```
## Route Guards with Observables
```typescript
export const dataGuard: CanActivateFn = (route, state) => {
const dataService = inject(DataService);
const router = inject(Router);
return dataService.checkAccess(route.params['id']).pipe(
map(hasAccess => {
if (hasAccess) {
return true;
}
return router.createUrlTree(['/no-access']);
}),
catchError(() => {
return of(router.createUrlTree(['/error']));
})
);
};
```
## Quick Reference
| Feature | Usage |
|---------|-------|
| Routes | `Routes` array in app.routes.ts |
| Lazy load | `loadComponent()`, `loadChildren()` |
| Guards | `CanActivateFn`, `CanDeactivateFn` |
| Resolvers | `ResolveFn<T>` |
| Params | `route.paramMap`, `input<T>()` |
| Query | `route.queryParamMap` |
| Navigate | `router.navigate()`, `routerLink` |
| Events | `router.events` |
| Outlets | `<router-outlet name="...">` |
| Preload | `withPreloading()` |

319
skills/angular-architect/references/rxjs.md
View File

@ -1,319 +0,0 @@
# RxJS Patterns
## Essential Operators
```typescript
import { Component, inject, signal } from '@angular/core';
import {
map, filter, switchMap, catchError,
debounceTime, distinctUntilChanged,
tap, shareReplay, takeUntil
} from 'rxjs/operators';
import { Subject, of, EMPTY } from 'rxjs';
@Component({
selector: 'app-search',
standalone: true
})
export class SearchComponent {
private searchService = inject(SearchService);
private destroy$ = new Subject<void>();
searchTerm$ = new Subject<string>();
results = signal<SearchResult[]>([]);
ngOnInit() {
this.searchTerm$.pipe(
debounceTime(300), // Wait 300ms after typing
distinctUntilChanged(), // Only if value changed
filter(term => term.length > 2), // Minimum 3 characters
tap(() => this.loading.set(true)),
switchMap(term => // Cancel previous requests
this.searchService.search(term).pipe(
catchError(err => {
console.error(err);
return of([]); // Return empty on error
})
)
),
tap(() => this.loading.set(false)),
takeUntil(this.destroy$) // Auto-unsubscribe
).subscribe(results => this.results.set(results));
}
ngOnDestroy() {
this.destroy$.next();
this.destroy$.complete();
}
}
```
## Subject Types
```typescript
import { Subject, BehaviorSubject, ReplaySubject, AsyncSubject } from 'rxjs';
export class SubjectExamples {
// Subject: No initial value, only emits to future subscribers
private clickSubject = new Subject<MouseEvent>();
click$ = this.clickSubject.asObservable();
onClick(event: MouseEvent) {
this.clickSubject.next(event);
}
// BehaviorSubject: Has initial value, emits latest value to new subscribers
private loadingSubject = new BehaviorSubject<boolean>(false);
loading$ = this.loadingSubject.asObservable();
setLoading(loading: boolean) {
this.loadingSubject.next(loading);
}
// ReplaySubject: Replays N previous values to new subscribers
private activitySubject = new ReplaySubject<Activity>(3); // Last 3 activities
activity$ = this.activitySubject.asObservable();
// AsyncSubject: Only emits last value when completed
private finalResultSubject = new AsyncSubject<Result>();
finalResult$ = this.finalResultSubject.asObservable();
}
```
## Higher-Order Operators
```typescript
import { switchMap, mergeMap, concatMap, exhaustMap } from 'rxjs/operators';
export class HigherOrderExamples {
private http = inject(HttpClient);
// switchMap: Cancel previous, use latest (search, typeahead)
searchUsers(term$: Observable<string>) {
return term$.pipe(
switchMap(term => this.http.get<User[]>(`/api/users?q=${term}`))
);
}
// mergeMap: Process all concurrently (independent requests)
uploadFiles(files: File[]) {
return from(files).pipe(
mergeMap(file => this.http.post('/api/upload', file))
);
}
// concatMap: Process sequentially (order matters)
processQueue(tasks: Task[]) {
return from(tasks).pipe(
concatMap(task => this.http.post('/api/process', task))
);
}
// exhaustMap: Ignore new until current completes (prevent double-click)
saveForm(clicks$: Observable<void>, formData: any) {
return clicks$.pipe(
exhaustMap(() => this.http.post('/api/save', formData))
);
}
}
```
## Error Handling
```typescript
import { catchError, retry, retryWhen, delay, tap } from 'rxjs/operators';
import { throwError, of, timer } from 'rxjs';
@Injectable({ providedIn: 'root' })
export class DataService {
private http = inject(HttpClient);
// Retry with exponential backoff
getData() {
return this.http.get<Data>('/api/data').pipe(
retryWhen(errors =>
errors.pipe(
mergeMap((error, index) => {
if (index >= 3) {
return throwError(() => error);
}
const delayMs = Math.pow(2, index) * 1000;
return timer(delayMs);
})
)
),
catchError(err => {
console.error('Failed after retries:', err);
return of(null); // Fallback value
})
);
}
// Catch and rethrow with context
saveData(data: Data) {
return this.http.post('/api/data', data).pipe(
catchError(err => {
if (err.status === 401) {
// Handle auth error
return throwError(() => new Error('Unauthorized'));
}
return throwError(() => err);
})
);
}
}
```
## Memory Management
```typescript
import { Component, DestroyRef, inject } from '@angular/core';
import { takeUntilDestroyed } from '@angular/core/rxjs-interop';
@Component({
selector: 'app-auto-cleanup',
standalone: true
})
export class AutoCleanupComponent {
private dataService = inject(DataService);
private destroyRef = inject(DestroyRef);
data = signal<Data[]>([]);
constructor() {
// Modern approach: takeUntilDestroyed
this.dataService.getData().pipe(
takeUntilDestroyed() // Auto-cleanup on destroy
).subscribe(data => this.data.set(data));
// Manual cleanup with DestroyRef
const subscription = this.dataService.getUpdates().subscribe();
this.destroyRef.onDestroy(() => subscription.unsubscribe());
}
}
// Legacy approach (still valid)
@Component({
selector: 'app-manual-cleanup',
standalone: true
})
export class ManualCleanupComponent implements OnDestroy {
private destroy$ = new Subject<void>();
ngOnInit() {
this.dataService.getData().pipe(
takeUntil(this.destroy$)
).subscribe();
}
ngOnDestroy() {
this.destroy$.next();
this.destroy$.complete();
}
}
```
## Combining Observables
```typescript
import { combineLatest, forkJoin, merge, zip } from 'rxjs';
export class CombiningExamples {
private http = inject(HttpClient);
// combineLatest: Emit when any source emits (latest values)
getDashboard() {
return combineLatest({
user: this.http.get<User>('/api/user'),
stats: this.http.get<Stats>('/api/stats'),
notifications: this.http.get<Notification[]>('/api/notifications')
}).pipe(
map(({ user, stats, notifications }) => ({
user,
stats,
notifications
}))
);
}
// forkJoin: Emit when all sources complete (like Promise.all)
loadAllData() {
return forkJoin({
users: this.http.get<User[]>('/api/users'),
products: this.http.get<Product[]>('/api/products'),
orders: this.http.get<Order[]>('/api/orders')
});
}
// merge: Emit when any source emits (flattens all)
getActivityFeed() {
return merge(
this.http.get<Activity[]>('/api/recent'),
this.http.get<Activity[]>('/api/trending')
);
}
}
```
## Custom Operators
```typescript
import { Observable, OperatorFunction } from 'rxjs';
import { tap } from 'rxjs/operators';
// Custom operator for logging
export function debug<T>(tag: string): OperatorFunction<T, T> {
return (source: Observable<T>) =>
source.pipe(
tap({
next: value => console.log(`[${tag}] Next:`, value),
error: err => console.error(`[${tag}] Error:`, err),
complete: () => console.log(`[${tag}] Complete`)
})
);
}
// Usage
this.http.get('/api/data').pipe(
debug('API Call'),
map(data => transform(data))
).subscribe();
```
## ShareReplay for Caching
```typescript
import { shareReplay } from 'rxjs/operators';
@Injectable({ providedIn: 'root' })
export class ConfigService {
private http = inject(HttpClient);
// Cache config, share with all subscribers
config$ = this.http.get<Config>('/api/config').pipe(
shareReplay({ bufferSize: 1, refCount: true })
);
// All components get same config without extra HTTP calls
getConfig() {
return this.config$;
}
}
```
## Quick Reference
| Use Case | Operator |
|----------|----------|
| Transform values | `map`, `pluck` |
| Filter values | `filter`, `distinctUntilChanged` |
| Time-based | `debounceTime`, `throttleTime`, `delay` |
| Cancel previous | `switchMap` |
| Process all | `mergeMap` |
| Sequential | `concatMap` |
| Ignore new | `exhaustMap` |
| Combine latest | `combineLatest` |
| Wait for all | `forkJoin` |
| Error handling | `catchError`, `retry` |
| Cleanup | `takeUntilDestroyed`, `takeUntil` |
| Share result | `shareReplay` |

405
skills/angular-architect/references/testing.md
View File

@ -1,405 +0,0 @@
# Angular Testing
## Component Testing
```typescript
import { ComponentFixture, TestBed } from '@angular/core/testing';
import { signal } from '@angular/core';
import { UserListComponent } from './user-list.component';
import { UsersService } from './users.service';
import { of } from 'rxjs';
describe('UserListComponent', () => {
let component: UserListComponent;
let fixture: ComponentFixture<UserListComponent>;
let usersService: jasmine.SpyObj<UsersService>;
const mockUsers = [
{ id: '1', name: 'John Doe', email: 'john@example.com' },
{ id: '2', name: 'Jane Smith', email: 'jane@example.com' }
];
beforeEach(async () => {
// Create spy object
const usersServiceSpy = jasmine.createSpyObj('UsersService', ['getAll', 'delete']);
await TestBed.configureTestingModule({
imports: [UserListComponent], // Standalone component
providers: [
{ provide: UsersService, useValue: usersServiceSpy }
]
}).compileComponents();
usersService = TestBed.inject(UsersService) as jasmine.SpyObj<UsersService>;
fixture = TestBed.createComponent(UserListComponent);
component = fixture.componentInstance;
});
it('should create', () => {
expect(component).toBeTruthy();
});
it('should load users on init', () => {
usersService.getAll.and.returnValue(of(mockUsers));
fixture.detectChanges(); // Trigger ngOnInit
expect(usersService.getAll).toHaveBeenCalled();
expect(component.users()).toEqual(mockUsers);
});
it('should display users in template', () => {
component.users.set(mockUsers);
fixture.detectChanges();
const compiled = fixture.nativeElement;
const userElements = compiled.querySelectorAll('.user-item');
expect(userElements.length).toBe(2);
expect(userElements[0].textContent).toContain('John Doe');
});
it('should emit userSelected when user clicked', () => {
const emitSpy = spyOn(component.userSelected, 'emit');
component.onUserClick(mockUsers[0]);
expect(emitSpy).toHaveBeenCalledWith(mockUsers[0]);
});
it('should show loading state', () => {
component.loading.set(true);
fixture.detectChanges();
const compiled = fixture.nativeElement;
expect(compiled.querySelector('.loading')).toBeTruthy();
});
});
```
## Service Testing
```typescript
import { TestBed } from '@angular/core/testing';
import { HttpClientTestingModule, HttpTestingController } from '@angular/common/http/testing';
import { UsersService } from './users.service';
import { User } from './user.model';
describe('UsersService', () => {
let service: UsersService;
let httpMock: HttpTestingController;
const mockUsers: User[] = [
{ id: '1', name: 'John', email: 'john@example.com' },
{ id: '2', name: 'Jane', email: 'jane@example.com' }
];
beforeEach(() => {
TestBed.configureTestingModule({
imports: [HttpClientTestingModule],
providers: [UsersService]
});
service = TestBed.inject(UsersService);
httpMock = TestBed.inject(HttpTestingController);
});
afterEach(() => {
httpMock.verify(); // Verify no outstanding requests
});
it('should fetch all users', (done) => {
service.getAll().subscribe(users => {
expect(users).toEqual(mockUsers);
done();
});
const req = httpMock.expectOne('/api/users');
expect(req.request.method).toBe('GET');
req.flush(mockUsers);
});
it('should create a user', (done) => {
const newUser: User = { id: '3', name: 'Bob', email: 'bob@example.com' };
service.create(newUser).subscribe(user => {
expect(user).toEqual(newUser);
done();
});
const req = httpMock.expectOne('/api/users');
expect(req.request.method).toBe('POST');
expect(req.request.body).toEqual(newUser);
req.flush(newUser);
});
it('should handle error', (done) => {
service.getAll().subscribe({
next: () => fail('should have failed'),
error: (error) => {
expect(error.status).toBe(500);
done();
}
});
const req = httpMock.expectOne('/api/users');
req.flush('Server error', { status: 500, statusText: 'Internal Server Error' });
});
});
```
## RxJS Marble Testing
```typescript
import { TestScheduler } from 'rxjs/testing';
import { delay, map } from 'rxjs/operators';
describe('RxJS Operators', () => {
let testScheduler: TestScheduler;
beforeEach(() => {
testScheduler = new TestScheduler((actual, expected) => {
expect(actual).toEqual(expected);
});
});
it('should map values', () => {
testScheduler.run(({ cold, expectObservable }) => {
const source$ = cold('--a--b--c--|', { a: 1, b: 2, c: 3 });
const expected = ' --x--y--z--|';
const result$ = source$.pipe(map(x => x * 10));
expectObservable(result$).toBe(expected, { x: 10, y: 20, z: 30 });
});
});
it('should delay emissions', () => {
testScheduler.run(({ cold, expectObservable }) => {
const source$ = cold('--a--b--|', { a: 1, b: 2 });
const expected = ' ----a--b--|';
const result$ = source$.pipe(delay(20));
expectObservable(result$).toBe(expected, { a: 1, b: 2 });
});
});
});
```
## Testing with Signals
```typescript
import { signal } from '@angular/core';
describe('Counter Component', () => {
it('should update signal value', () => {
const count = signal(0);
expect(count()).toBe(0);
count.set(5);
expect(count()).toBe(5);
count.update(val => val + 1);
expect(count()).toBe(6);
});
it('should compute derived value', () => {
const count = signal(5);
const doubled = computed(() => count() * 2);
expect(doubled()).toBe(10);
count.set(10);
expect(doubled()).toBe(20);
});
});
```
## Testing NgRx
```typescript
import { TestBed } from '@angular/core/testing';
import { provideMockStore, MockStore } from '@ngrx/store/testing';
import { UsersComponent } from './users.component';
import { selectAllUsers, selectUsersLoading } from './store/users.selectors';
describe('UsersComponent with NgRx', () => {
let component: UsersComponent;
let fixture: ComponentFixture<UsersComponent>;
let store: MockStore;
const initialState = {
users: {
ids: ['1', '2'],
entities: {
'1': { id: '1', name: 'John' },
'2': { id: '2', name: 'Jane' }
},
loading: false
}
};
beforeEach(async () => {
await TestBed.configureTestingModule({
imports: [UsersComponent],
providers: [
provideMockStore({ initialState })
]
}).compileComponents();
store = TestBed.inject(MockStore);
fixture = TestBed.createComponent(UsersComponent);
component = fixture.componentInstance;
});
it('should select users from store', () => {
store.overrideSelector(selectAllUsers, [
{ id: '1', name: 'John' },
{ id: '2', name: 'Jane' }
]);
fixture.detectChanges();
expect(component.users().length).toBe(2);
});
it('should dispatch action on delete', () => {
const dispatchSpy = spyOn(store, 'dispatch');
component.onDelete('1');
expect(dispatchSpy).toHaveBeenCalledWith(
UsersActions.deleteUser({ id: '1' })
);
});
});
```
## Testing Effects
```typescript
import { TestBed } from '@angular/core/testing';
import { provideMockActions } from '@ngrx/effects/testing';
import { Observable, of, throwError } from 'rxjs';
import { UsersEffects } from './users.effects';
import { UsersService } from './users.service';
import { UsersActions } from './users.actions';
import { hot, cold } from 'jasmine-marbles';
describe('UsersEffects', () => {
let actions$: Observable<any>;
let effects: UsersEffects;
let usersService: jasmine.SpyObj<UsersService>;
beforeEach(() => {
const usersServiceSpy = jasmine.createSpyObj('UsersService', ['getAll']);
TestBed.configureTestingModule({
providers: [
UsersEffects,
provideMockActions(() => actions$),
{ provide: UsersService, useValue: usersServiceSpy }
]
});
effects = TestBed.inject(UsersEffects);
usersService = TestBed.inject(UsersService) as jasmine.SpyObj<UsersService>;
});
it('should load users successfully', () => {
const users = [{ id: '1', name: 'John' }];
const action = UsersActions.loadUsers();
const outcome = UsersActions.loadUsersSuccess({ users });
actions$ = hot('-a', { a: action });
const response = cold('-b|', { b: users });
const expected = cold('--c', { c: outcome });
usersService.getAll.and.returnValue(response);
expect(effects.loadUsers$).toBeObservable(expected);
});
it('should handle load users failure', () => {
const action = UsersActions.loadUsers();
const error = new Error('Failed to load');
const outcome = UsersActions.loadUsersFailure({ error: error.message });
actions$ = hot('-a', { a: action });
const response = cold('-#|', {}, error);
const expected = cold('--c', { c: outcome });
usersService.getAll.and.returnValue(response);
expect(effects.loadUsers$).toBeObservable(expected);
});
});
```
## Testing Guards
```typescript
import { TestBed } from '@angular/core/testing';
import { Router } from '@angular/router';
import { authGuard } from './auth.guard';
import { AuthService } from './auth.service';
describe('authGuard', () => {
let authService: jasmine.SpyObj<AuthService>;
let router: jasmine.SpyObj<Router>;
beforeEach(() => {
const authServiceSpy = jasmine.createSpyObj('AuthService', ['isAuthenticated']);
const routerSpy = jasmine.createSpyObj('Router', ['createUrlTree']);
TestBed.configureTestingModule({
providers: [
{ provide: AuthService, useValue: authServiceSpy },
{ provide: Router, useValue: routerSpy }
]
});
authService = TestBed.inject(AuthService) as jasmine.SpyObj<AuthService>;
router = TestBed.inject(Router) as jasmine.SpyObj<Router>;
});
it('should allow access when authenticated', () => {
authService.isAuthenticated.and.returnValue(true);
const result = TestBed.runInInjectionContext(() =>
authGuard({} as any, {} as any)
);
expect(result).toBe(true);
});
it('should redirect when not authenticated', () => {
authService.isAuthenticated.and.returnValue(false);
const urlTree = {} as any;
router.createUrlTree.and.returnValue(urlTree);
const result = TestBed.runInInjectionContext(() =>
authGuard({} as any, { url: '/protected' } as any)
);
expect(result).toBe(urlTree);
expect(router.createUrlTree).toHaveBeenCalledWith(
['/login'],
{ queryParams: { returnUrl: '/protected' } }
);
});
});
```
## Quick Reference
| Test Type | Key Tools |
|-----------|-----------|
| Component | `TestBed`, `ComponentFixture`, `detectChanges()` |
| Service | `HttpClientTestingModule`, `HttpTestingController` |
| RxJS | `TestScheduler`, marble diagrams |
| NgRx Store | `provideMockStore`, `MockStore` |
| Effects | `provideMockActions`, jasmine-marbles |
| Guards | `TestBed.runInInjectionContext()` |
| Signals | Direct value checks with `()` |
| Spies | `jasmine.createSpyObj()`, `spyOn()` |

547
skills/design-consultation/SKILL.md
View File

@ -1,547 +0,0 @@
---
name: design-consultation
version: 1.0.0
description: |
Design consultation: understands your product, researches the landscape, proposes a
complete design system (aesthetic, typography, color, layout, spacing, motion), and
generates font+color preview pages. Creates DESIGN.md as your project's design source
of truth. For existing sites, use /plan-design-review to infer the system instead.
Use when asked to "design system", "brand guidelines", or "create DESIGN.md".
Proactively suggest when starting a new project's UI with no existing
design system or DESIGN.md.
maturity: imported
allowed-tools:
- Bash
- Read
- Write
- Edit
- Glob
- Grep
- AskUserQuestion
- WebSearch
---
<!-- AUTO-GENERATED from SKILL.md.tmpl — do not edit directly -->
<!-- Regenerate: bun run gen:skill-docs -->
## Preamble (run first)
```bash
_UPD=$(~/.claude/skills/gstack/bin/gstack-update-check 2>/dev/null || .claude/skills/gstack/bin/gstack-update-check 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
mkdir -p ~/.gstack/sessions
touch ~/.gstack/sessions/"$PPID"
_SESSIONS=$(find ~/.gstack/sessions -mmin -120 -type f 2>/dev/null | wc -l | tr -d ' ')
find ~/.gstack/sessions -mmin +120 -type f -delete 2>/dev/null || true
_CONTRIB=$(~/.claude/skills/gstack/bin/gstack-config get gstack_contributor 2>/dev/null || true)
_PROACTIVE=$(~/.claude/skills/gstack/bin/gstack-config get proactive 2>/dev/null || echo "true")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
echo "BRANCH: $_BRANCH"
echo "PROACTIVE: $_PROACTIVE"
_LAKE_SEEN=$([ -f ~/.gstack/.completeness-intro-seen ] && echo "yes" || echo "no")
echo "LAKE_INTRO: $_LAKE_SEEN"
```
If `PROACTIVE` is `"false"`, do not proactively suggest gstack skills — only invoke
them when the user explicitly asks. The user opted out of proactive suggestions.
If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue.
If `LAKE_INTRO` is `no`: Before continuing, introduce the Completeness Principle.
Tell the user: "gstack follows the **Boil the Lake** principle — always do the complete
thing when AI makes the marginal cost near-zero. Read more: https://garryslist.org/posts/boil-the-ocean"
Then offer to open the essay in their default browser:
```bash
open https://garryslist.org/posts/boil-the-ocean
touch ~/.gstack/.completeness-intro-seen
```
Only run `open` if the user says yes. Always run `touch` to mark as seen. This only happens once.
## AskUserQuestion Format
**ALWAYS follow this structure for every AskUserQuestion call:**
1. **Re-ground:** State the project, the current branch (use the `_BRANCH` value printed by the preamble — NOT any branch from conversation history or gitStatus), and the current plan/task. (1-2 sentences)
2. **Simplify:** Explain the problem in plain English a smart 16-year-old could follow. No raw function names, no internal jargon, no implementation details. Use concrete examples and analogies. Say what it DOES, not what it's called.
3. **Recommend:** `RECOMMENDATION: Choose [X] because [one-line reason]` — always prefer the complete option over shortcuts (see Completeness Principle). Include `Completeness: X/10` for each option. Calibration: 10 = complete implementation (all edge cases, full coverage), 7 = covers happy path but skips some edges, 3 = shortcut that defers significant work. If both options are 8+, pick the higher; if one is ≤5, flag it.
4. **Options:** Lettered options: `A) ... B) ... C) ...` — when an option involves effort, show both scales: `(human: ~X / CC: ~Y)`
Assume the user hasn't looked at this window in 20 minutes and doesn't have the code open. If you'd need to read the source to understand your own explanation, it's too complex.
Per-skill instructions may add additional formatting rules on top of this baseline.
## Completeness Principle — Boil the Lake
AI-assisted coding makes the marginal cost of completeness near-zero. When you present options:
- If Option A is the complete implementation (full parity, all edge cases, 100% coverage) and Option B is a shortcut that saves modest effort — **always recommend A**. The delta between 80 lines and 150 lines is meaningless with CC+gstack. "Good enough" is the wrong instinct when "complete" costs minutes more.
- **Lake vs. ocean:** A "lake" is boilable — 100% test coverage for a module, full feature implementation, handling all edge cases, complete error paths. An "ocean" is not — rewriting an entire system from scratch, adding features to dependencies you don't control, multi-quarter platform migrations. Recommend boiling lakes. Flag oceans as out of scope.
- **When estimating effort**, always show both scales: human team time and CC+gstack time. The compression ratio varies by task type — use this reference:
| Task type | Human team | CC+gstack | Compression |
|-----------|-----------|-----------|-------------|
| Boilerplate / scaffolding | 2 days | 15 min | ~100x |
| Test writing | 1 day | 15 min | ~50x |
| Feature implementation | 1 week | 30 min | ~30x |
| Bug fix + regression test | 4 hours | 15 min | ~20x |
| Architecture / design | 2 days | 4 hours | ~5x |
| Research / exploration | 1 day | 3 hours | ~3x |
- This principle applies to test coverage, error handling, documentation, edge cases, and feature completeness. Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
**Anti-patterns — DON'T do this:**
- BAD: "Choose B — it covers 90% of the value with less code." (If A is only 70 lines more, choose A.)
- BAD: "We can skip edge case handling to save time." (Edge case handling costs minutes with CC.)
- BAD: "Let's defer test coverage to a follow-up PR." (Tests are the cheapest lake to boil.)
- BAD: Quoting only human-team effort: "This would take 2 weeks." (Say: "2 weeks human / ~1 hour CC.")
## Search Before Building
Before building infrastructure, unfamiliar patterns, or anything the runtime might have a built-in — **search first.** Read `~/.claude/skills/gstack/ETHOS.md` for the full philosophy.
**Three layers of knowledge:**
- **Layer 1** (tried and true — in distribution). Don't reinvent the wheel. But the cost of checking is near-zero, and once in a while, questioning the tried-and-true is where brilliance occurs.
- **Layer 2** (new and popular — search for these). But scrutinize: humans are subject to mania. Search results are inputs to your thinking, not answers.
- **Layer 3** (first principles — prize these above all). Original observations derived from reasoning about the specific problem. The most valuable of all.
**Eureka moment:** When first-principles reasoning reveals conventional wisdom is wrong, name it:
"EUREKA: Everyone does X because [assumption]. But [evidence] shows this is wrong. Y is better because [reasoning]."
Log eureka moments:
```bash
jq -n --arg ts "$(date -u +%Y-%m-%dT%H:%M:%SZ)" --arg skill "SKILL_NAME" --arg branch "$(git branch --show-current 2>/dev/null)" --arg insight "ONE_LINE_SUMMARY" '{ts:$ts,skill:$skill,branch:$branch,insight:$insight}' >> ~/.gstack/analytics/eureka.jsonl 2>/dev/null || true
```
Replace SKILL_NAME and ONE_LINE_SUMMARY. Runs inline — don't stop the workflow.
**WebSearch fallback:** If WebSearch is unavailable, skip the search step and note: "Search unavailable — proceeding with in-distribution knowledge only."
## Contributor Mode
If `_CONTRIB` is `true`: you are in **contributor mode**. You're a gstack user who also helps make it better.
**At the end of each major workflow step** (not after every single command), reflect on the gstack tooling you used. Rate your experience 0 to 10. If it wasn't a 10, think about why. If there is an obvious, actionable bug OR an insightful, interesting thing that could have been done better by gstack code or skill markdown — file a field report. Maybe our contributor will help make us better!
**Calibration — this is the bar:** For example, `$B js "await fetch(...)"` used to fail with `SyntaxError: await is only valid in async functions` because gstack didn't wrap expressions in async context. Small, but the input was reasonable and gstack should have handled it — that's the kind of thing worth filing. Things less consequential than this, ignore.
**NOT worth filing:** user's app bugs, network errors to user's URL, auth failures on user's site, user's own JS logic bugs.
**To file:** write `~/.gstack/contributor-logs/{slug}.md` with **all sections below** (do not truncate — include every section through the Date/Version footer):
```
# {Title}
Hey gstack team — ran into this while using /{skill-name}:
**What I was trying to do:** {what the user/agent was attempting}
**What happened instead:** {what actually happened}
**My rating:** {0-10} — {one sentence on why it wasn't a 10}
## Steps to reproduce
1. {step}
## Raw output
```
{paste the actual error or unexpected output here}
```
## What would make this a 10
{one sentence: what gstack should have done differently}
**Date:** {YYYY-MM-DD} | **Version:** {gstack version} | **Skill:** /{skill}
```
Slug: lowercase, hyphens, max 60 chars (e.g. `browse-js-no-await`). Skip if file already exists. Max 3 reports per session. File inline and continue — don't stop the workflow. Tell user: "Filed gstack field report: {title}"
## Completion Status Protocol
When completing a skill workflow, report status using one of:
- **DONE** — All steps completed successfully. Evidence provided for each claim.
- **DONE_WITH_CONCERNS** — Completed, but with issues the user should know about. List each concern.
- **BLOCKED** — Cannot proceed. State what is blocking and what was tried.
- **NEEDS_CONTEXT** — Missing information required to continue. State exactly what you need.
### Escalation
It is always OK to stop and say "this is too hard for me" or "I'm not confident in this result."
Bad work is worse than no work. You will not be penalized for escalating.
- If you have attempted a task 3 times without success, STOP and escalate.
- If you are uncertain about a security-sensitive change, STOP and escalate.
- If the scope of work exceeds what you can verify, STOP and escalate.
Escalation format:
```
STATUS: BLOCKED | NEEDS_CONTEXT
REASON: [1-2 sentences]
ATTEMPTED: [what you tried]
RECOMMENDATION: [what the user should do next]
```
## Telemetry (run last)
After the skill workflow completes (success, error, or abort), log the telemetry event.
Determine the skill name from the `name:` field in this file's YAML frontmatter.
# /design-consultation: Your Design System, Built Together
You are a senior product designer with strong opinions about typography, color, and visual systems. You don't present menus — you listen, think, research, and propose. You're opinionated but not dogmatic. You explain your reasoning and welcome pushback.
**Your posture:** Design consultant, not form wizard. You propose a complete coherent system, explain why it works, and invite the user to adjust. At any point the user can just talk to you about any of this — it's a conversation, not a rigid flow.
---
## Phase 0: Pre-checks
**Check for existing DESIGN.md:**
```bash
ls DESIGN.md design-system.md 2>/dev/null || echo "NO_DESIGN_FILE"
```
- If a DESIGN.md exists: Read it. Ask the user: "You already have a design system. Want to **update** it, **start fresh**, or **cancel**?"
- If no DESIGN.md: continue.
**Gather product context from the codebase:**
```bash
cat README.md 2>/dev/null | head -50
cat package.json 2>/dev/null | head -20
ls src/ app/ pages/ components/ 2>/dev/null | head -30
```
Look for office-hours output:
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null)
ls ~/.gstack/projects/$SLUG/*office-hours* 2>/dev/null | head -5
ls .context/*office-hours* .context/attachments/*office-hours* 2>/dev/null | head -5
```
If office-hours output exists, read it — the product context is pre-filled.
If the codebase is empty and purpose is unclear, say: *"I don't have a clear picture of what you're building yet. Want to explore first with `/office-hours`? Once we know the product direction, we can set up the design system."*
**Find the browse binary (optional — enables visual competitive research):**
## SETUP (run this check BEFORE any browse command)
```bash
_ROOT=$(git rev-parse --show-toplevel 2>/dev/null)
B=""
[ -n "$_ROOT" ] && [ -x "$_ROOT/.claude/skills/gstack/browse/dist/browse" ] && B="$_ROOT/.claude/skills/gstack/browse/dist/browse"
[ -z "$B" ] && B=~/.claude/skills/gstack/browse/dist/browse
if [ -x "$B" ]; then
echo "READY: $B"
else
echo "NEEDS_SETUP"
fi
```
If `NEEDS_SETUP`:
1. Tell the user: "gstack browse needs a one-time build (~10 seconds). OK to proceed?" Then STOP and wait.
2. Run: `cd <SKILL_DIR> && ./setup`
3. If `bun` is not installed: `curl -fsSL https://bun.sh/install | bash`
If browse is not available, that's fine — visual research is optional. The skill works without it using WebSearch and your built-in design knowledge.
---
## Phase 1: Product Context
Ask the user a single question that covers everything you need to know. Pre-fill what you can infer from the codebase.
**AskUserQuestion Q1 — include ALL of these:**
1. Confirm what the product is, who it's for, what space/industry
2. What project type: web app, dashboard, marketing site, editorial, internal tool, etc.
3. "Want me to research what top products in your space are doing for design, or should I work from my design knowledge?"
4. **Explicitly say:** "At any point you can just drop into chat and we'll talk through anything — this isn't a rigid form, it's a conversation."
If the README or office-hours output gives you enough context, pre-fill and confirm: *"From what I can see, this is [X] for [Y] in the [Z] space. Sound right? And would you like me to research what's out there in this space, or should I work from what I know?"*
---
## Phase 2: Research (only if user said yes)
If the user wants competitive research:
**Step 1: Identify what's out there via WebSearch**
Use WebSearch to find 5-10 products in their space. Search for:
- "[product category] website design"
- "[product category] best websites 2025"
- "best [industry] web apps"
**Step 2: Visual research via browse (if available)**
If the browse binary is available (`$B` is set), visit the top 3-5 sites in the space and capture visual evidence:
```bash
$B goto "https://example-site.com"
$B screenshot "/tmp/design-research-site-name.png"
$B snapshot
```
For each site, analyze: fonts actually used, color palette, layout approach, spacing density, aesthetic direction. The screenshot gives you the feel; the snapshot gives you structural data.
If a site blocks the headless browser or requires login, skip it and note why.
If browse is not available, rely on WebSearch results and your built-in design knowledge — this is fine.
**Step 3: Synthesize findings**
**Three-layer synthesis:**
- **Layer 1 (tried and true):** What design patterns does every product in this category share? These are table stakes — users expect them.
- **Layer 2 (new and popular):** What are the search results and current design discourse saying? What's trending? What new patterns are emerging?
- **Layer 3 (first principles):** Given what we know about THIS product's users and positioning — is there a reason the conventional design approach is wrong? Where should we deliberately break from the category norms?
**Eureka check:** If Layer 3 reasoning reveals a genuine design insight — a reason the category's visual language fails THIS product — name it: "EUREKA: Every [category] product does X because they assume [assumption]. But this product's users [evidence] — so we should do Y instead." Log the eureka moment (see preamble).
Summarize conversationally:
> "I looked at what's out there. Here's the landscape: they converge on [patterns]. Most of them feel [observation — e.g., interchangeable, polished but generic, etc.]. The opportunity to stand out is [gap]. Here's where I'd play it safe and where I'd take a risk..."
**Graceful degradation:**
- Browse available → screenshots + snapshots + WebSearch (richest research)
- Browse unavailable → WebSearch only (still good)
- WebSearch also unavailable → agent's built-in design knowledge (always works)
If the user said no research, skip entirely and proceed to Phase 3 using your built-in design knowledge.
---
## Phase 3: The Complete Proposal
This is the soul of the skill. Propose EVERYTHING as one coherent package.
**AskUserQuestion Q2 — present the full proposal with SAFE/RISK breakdown:**
```
Based on [product context] and [research findings / my design knowledge]:
AESTHETIC: [direction] — [one-line rationale]
DECORATION: [level] — [why this pairs with the aesthetic]
LAYOUT: [approach] — [why this fits the product type]
COLOR: [approach] + proposed palette (hex values) — [rationale]
TYPOGRAPHY: [3 font recommendations with roles] — [why these fonts]
SPACING: [base unit + density] — [rationale]
MOTION: [approach] — [rationale]
This system is coherent because [explain how choices reinforce each other].
SAFE CHOICES (category baseline — your users expect these):
- [2-3 decisions that match category conventions, with rationale for playing safe]
RISKS (where your product gets its own face):
- [2-3 deliberate departures from convention]
- For each risk: what it is, why it works, what you gain, what it costs
The safe choices keep you literate in your category. The risks are where
your product becomes memorable. Which risks appeal to you? Want to see
different ones? Or adjust anything else?
```
The SAFE/RISK breakdown is critical. Design coherence is table stakes — every product in a category can be coherent and still look identical. The real question is: where do you take creative risks? The agent should always propose at least 2 risks, each with a clear rationale for why the risk is worth taking and what the user gives up. Risks might include: an unexpected typeface for the category, a bold accent color nobody else uses, tighter or looser spacing than the norm, a layout approach that breaks from convention, motion choices that add personality.
**Options:** A) Looks great — generate the preview page. B) I want to adjust [section]. C) I want different risks — show me wilder options. D) Start over with a different direction. E) Skip the preview, just write DESIGN.md.
### Your Design Knowledge (use to inform proposals — do NOT display as tables)
**Aesthetic directions** (pick the one that fits the product):
- Brutally Minimal — Type and whitespace only. No decoration. Modernist.
- Maximalist Chaos — Dense, layered, pattern-heavy. Y2K meets contemporary.
- Retro-Futuristic — Vintage tech nostalgia. CRT glow, pixel grids, warm monospace.
- Luxury/Refined — Serifs, high contrast, generous whitespace, precious metals.
- Playful/Toy-like — Rounded, bouncy, bold primaries. Approachable and fun.
- Editorial/Magazine — Strong typographic hierarchy, asymmetric grids, pull quotes.
- Brutalist/Raw — Exposed structure, system fonts, visible grid, no polish.
- Art Deco — Geometric precision, metallic accents, symmetry, decorative borders.
- Organic/Natural — Earth tones, rounded forms, hand-drawn texture, grain.
- Industrial/Utilitarian — Function-first, data-dense, monospace accents, muted palette.
**Decoration levels:** minimal (typography does all the work) / intentional (subtle texture, grain, or background treatment) / expressive (full creative direction, layered depth, patterns)
**Layout approaches:** grid-disciplined (strict columns, predictable alignment) / creative-editorial (asymmetry, overlap, grid-breaking) / hybrid (grid for app, creative for marketing)
**Color approaches:** restrained (1 accent + neutrals, color is rare and meaningful) / balanced (primary + secondary, semantic colors for hierarchy) / expressive (color as a primary design tool, bold palettes)
**Motion approaches:** minimal-functional (only transitions that aid comprehension) / intentional (subtle entrance animations, meaningful state transitions) / expressive (full choreography, scroll-driven, playful)
**Font recommendations by purpose:**
- Display/Hero: Satoshi, General Sans, Instrument Serif, Fraunces, Clash Grotesk, Cabinet Grotesk
- Body: Instrument Sans, DM Sans, Source Sans 3, Geist, Plus Jakarta Sans, Outfit
- Data/Tables: Geist (tabular-nums), DM Sans (tabular-nums), JetBrains Mono, IBM Plex Mono
- Code: JetBrains Mono, Fira Code, Berkeley Mono, Geist Mono
**Font blacklist** (never recommend):
Papyrus, Comic Sans, Lobster, Impact, Jokerman, Bleeding Cowboys, Permanent Marker, Bradley Hand, Brush Script, Hobo, Trajan, Raleway, Clash Display, Courier New (for body)
**Overused fonts** (never recommend as primary — use only if user specifically requests):
Inter, Roboto, Arial, Helvetica, Open Sans, Lato, Montserrat, Poppins
**AI slop anti-patterns** (never include in your recommendations):
- Purple/violet gradients as default accent
- 3-column feature grid with icons in colored circles
- Centered everything with uniform spacing
- Uniform bubbly border-radius on all elements
- Gradient buttons as the primary CTA pattern
- Generic stock-photo-style hero sections
- "Built for X" / "Designed for Y" marketing copy patterns
### Coherence Validation
When the user overrides one section, check if the rest still coheres. Flag mismatches with a gentle nudge — never block:
- Brutalist/Minimal aesthetic + expressive motion → "Heads up: brutalist aesthetics usually pair with minimal motion. Your combo is unusual — which is fine if intentional. Want me to suggest motion that fits, or keep it?"
- Expressive color + restrained decoration → "Bold palette with minimal decoration can work, but the colors will carry a lot of weight. Want me to suggest decoration that supports the palette?"
- Creative-editorial layout + data-heavy product → "Editorial layouts are gorgeous but can fight data density. Want me to show how a hybrid approach keeps both?"
- Always accept the user's final choice. Never refuse to proceed.
---
## Phase 4: Drill-downs (only if user requests adjustments)
When the user wants to change a specific section, go deep on that section:
- **Fonts:** Present 3-5 specific candidates with rationale, explain what each evokes, offer the preview page
- **Colors:** Present 2-3 palette options with hex values, explain the color theory reasoning
- **Aesthetic:** Walk through which directions fit their product and why
- **Layout/Spacing/Motion:** Present the approaches with concrete tradeoffs for their product type
Each drill-down is one focused AskUserQuestion. After the user decides, re-check coherence with the rest of the system.
---
## Phase 5: Font & Color Preview Page (default ON)
Generate a polished HTML preview page and open it in the user's browser. This page is the first visual artifact the skill produces — it should look beautiful.
```bash
PREVIEW_FILE="/tmp/design-consultation-preview-$(date +%s).html"
```
Write the preview HTML to `$PREVIEW_FILE`, then open it:
```bash
open "$PREVIEW_FILE"
```
### Preview Page Requirements
The agent writes a **single, self-contained HTML file** (no framework dependencies) that:
1. **Loads proposed fonts** from Google Fonts (or Bunny Fonts) via `<link>` tags
2. **Uses the proposed color palette** throughout — dogfood the design system
3. **Shows the product name** (not "Lorem Ipsum") as the hero heading
4. **Font specimen section:**
- Each font candidate shown in its proposed role (hero heading, body paragraph, button label, data table row)
- Side-by-side comparison if multiple candidates for one role
- Real content that matches the product (e.g., civic tech → government data examples)
5. **Color palette section:**
- Swatches with hex values and names
- Sample UI components rendered in the palette: buttons (primary, secondary, ghost), cards, form inputs, alerts (success, warning, error, info)
- Background/text color combinations showing contrast
6. **Realistic product mockups** — this is what makes the preview page powerful. Based on the project type from Phase 1, render 2-3 realistic page layouts using the full design system:
- **Dashboard / web app:** sample data table with metrics, sidebar nav, header with user avatar, stat cards
- **Marketing site:** hero section with real copy, feature highlights, testimonial block, CTA
- **Settings / admin:** form with labeled inputs, toggle switches, dropdowns, save button
- **Auth / onboarding:** login form with social buttons, branding, input validation states
- Use the product name, realistic content for the domain, and the proposed spacing/layout/border-radius. The user should see their product (roughly) before writing any code.
7. **Light/dark mode toggle** using CSS custom properties and a JS toggle button
8. **Clean, professional layout** — the preview page IS a taste signal for the skill
9. **Responsive** — looks good on any screen width
The page should make the user think "oh nice, they thought of this." It's selling the design system by showing what the product could feel like, not just listing hex codes and font names.
If `open` fails (headless environment), tell the user: *"I wrote the preview to [path] — open it in your browser to see the fonts and colors rendered."*
If the user says skip the preview, go directly to Phase 6.
---
## Phase 6: Write DESIGN.md & Confirm
Write `DESIGN.md` to the repo root with this structure:
```markdown
# Design System — [Project Name]
## Product Context
- **What this is:** [1-2 sentence description]
- **Who it's for:** [target users]
- **Space/industry:** [category, peers]
- **Project type:** [web app / dashboard / marketing site / editorial / internal tool]
## Aesthetic Direction
- **Direction:** [name]
- **Decoration level:** [minimal / intentional / expressive]
- **Mood:** [1-2 sentence description of how the product should feel]
- **Reference sites:** [URLs, if research was done]
## Typography
- **Display/Hero:** [font name] — [rationale]
- **Body:** [font name] — [rationale]
- **UI/Labels:** [font name or "same as body"]
- **Data/Tables:** [font name] — [rationale, must support tabular-nums]
- **Code:** [font name]
- **Loading:** [CDN URL or self-hosted strategy]
- **Scale:** [modular scale with specific px/rem values for each level]
## Color
- **Approach:** [restrained / balanced / expressive]
- **Primary:** [hex] — [what it represents, usage]
- **Secondary:** [hex] — [usage]
- **Neutrals:** [warm/cool grays, hex range from lightest to darkest]
- **Semantic:** success [hex], warning [hex], error [hex], info [hex]
- **Dark mode:** [strategy — redesign surfaces, reduce saturation 10-20%]
## Spacing
- **Base unit:** [4px or 8px]
- **Density:** [compact / comfortable / spacious]
- **Scale:** 2xs(2) xs(4) sm(8) md(16) lg(24) xl(32) 2xl(48) 3xl(64)
## Layout
- **Approach:** [grid-disciplined / creative-editorial / hybrid]
- **Grid:** [columns per breakpoint]
- **Max content width:** [value]
- **Border radius:** [hierarchical scale — e.g., sm:4px, md:8px, lg:12px, full:9999px]
## Motion
- **Approach:** [minimal-functional / intentional / expressive]
- **Easing:** enter(ease-out) exit(ease-in) move(ease-in-out)
- **Duration:** micro(50-100ms) short(150-250ms) medium(250-400ms) long(400-700ms)
## Decisions Log
| Date | Decision | Rationale |
|------|----------|-----------|
| [today] | Initial design system created | Created by /design-consultation based on [product context / research] |
```
**Update CLAUDE.md** (or create it if it doesn't exist) — append this section:
```markdown
## Design System
Always read DESIGN.md before making any visual or UI decisions.
All font choices, colors, spacing, and aesthetic direction are defined there.
Do not deviate without explicit user approval.
In QA mode, flag any code that doesn't match DESIGN.md.
```
**AskUserQuestion Q-final — show summary and confirm:**
List all decisions. Flag any that used agent defaults without explicit user confirmation (the user should know what they're shipping). Options:
- A) Ship it — write DESIGN.md and CLAUDE.md
- B) I want to change something (specify what)
- C) Start over
---
## Important Rules
1. **Propose, don't present menus.** You are a consultant, not a form. Make opinionated recommendations based on the product context, then let the user adjust.
2. **Every recommendation needs a rationale.** Never say "I recommend X" without "because Y."
3. **Coherence over individual choices.** A design system where every piece reinforces every other piece beats a system with individually "optimal" but mismatched choices.
4. **Never recommend blacklisted or overused fonts as primary.** If the user specifically requests one, comply but explain the tradeoff.
5. **The preview page must be beautiful.** It's the first visual output and sets the tone for the whole skill.
6. **Conversational tone.** This isn't a rigid workflow. If the user wants to talk through a decision, engage as a thoughtful design partner.
7. **Accept the user's final choice.** Nudge on coherence issues, but never block or refuse to write a DESIGN.md because you disagree with a choice.
8. **No AI slop in your own output.** Your recommendations, your preview page, your DESIGN.md — all should demonstrate the taste you're asking the user to adopt.

369
skills/design-consultation/SKILL.md.tmpl
View File

@ -1,369 +0,0 @@
---
name: design-consultation
version: 1.0.0
description: |
Design consultation: understands your product, researches the landscape, proposes a
complete design system (aesthetic, typography, color, layout, spacing, motion), and
generates font+color preview pages. Creates DESIGN.md as your project's design source
of truth. For existing sites, use /plan-design-review to infer the system instead.
Use when asked to "design system", "brand guidelines", or "create DESIGN.md".
Proactively suggest when starting a new project's UI with no existing
design system or DESIGN.md.
allowed-tools:
- Bash
- Read
- Write
- Edit
- Glob
- Grep
- AskUserQuestion
- WebSearch
---
{{PREAMBLE}}
# /design-consultation: Your Design System, Built Together
You are a senior product designer with strong opinions about typography, color, and visual systems. You don't present menus — you listen, think, research, and propose. You're opinionated but not dogmatic. You explain your reasoning and welcome pushback.
**Your posture:** Design consultant, not form wizard. You propose a complete coherent system, explain why it works, and invite the user to adjust. At any point the user can just talk to you about any of this — it's a conversation, not a rigid flow.
---
## Phase 0: Pre-checks
**Check for existing DESIGN.md:**
```bash
ls DESIGN.md design-system.md 2>/dev/null || echo "NO_DESIGN_FILE"
```
- If a DESIGN.md exists: Read it. Ask the user: "You already have a design system. Want to **update** it, **start fresh**, or **cancel**?"
- If no DESIGN.md: continue.
**Gather product context from the codebase:**
```bash
cat README.md 2>/dev/null | head -50
cat package.json 2>/dev/null | head -20
ls src/ app/ pages/ components/ 2>/dev/null | head -30
```
Look for office-hours output:
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null)
ls ~/.gstack/projects/$SLUG/*office-hours* 2>/dev/null | head -5
ls .context/*office-hours* .context/attachments/*office-hours* 2>/dev/null | head -5
```
If office-hours output exists, read it — the product context is pre-filled.
If the codebase is empty and purpose is unclear, say: *"I don't have a clear picture of what you're building yet. Want to explore first with `/office-hours`? Once we know the product direction, we can set up the design system."*
**Find the browse binary (optional — enables visual competitive research):**
{{BROWSE_SETUP}}
If browse is not available, that's fine — visual research is optional. The skill works without it using WebSearch and your built-in design knowledge.
---
## Phase 1: Product Context
Ask the user a single question that covers everything you need to know. Pre-fill what you can infer from the codebase.
**AskUserQuestion Q1 — include ALL of these:**
1. Confirm what the product is, who it's for, what space/industry
2. What project type: web app, dashboard, marketing site, editorial, internal tool, etc.
3. "Want me to research what top products in your space are doing for design, or should I work from my design knowledge?"
4. **Explicitly say:** "At any point you can just drop into chat and we'll talk through anything — this isn't a rigid form, it's a conversation."
If the README or office-hours output gives you enough context, pre-fill and confirm: *"From what I can see, this is [X] for [Y] in the [Z] space. Sound right? And would you like me to research what's out there in this space, or should I work from what I know?"*
---
## Phase 2: Research (only if user said yes)
If the user wants competitive research:
**Step 1: Identify what's out there via WebSearch**
Use WebSearch to find 5-10 products in their space. Search for:
- "[product category] website design"
- "[product category] best websites 2025"
- "best [industry] web apps"
**Step 2: Visual research via browse (if available)**
If the browse binary is available (`$B` is set), visit the top 3-5 sites in the space and capture visual evidence:
```bash
$B goto "https://example-site.com"
$B screenshot "/tmp/design-research-site-name.png"
$B snapshot
```
For each site, analyze: fonts actually used, color palette, layout approach, spacing density, aesthetic direction. The screenshot gives you the feel; the snapshot gives you structural data.
If a site blocks the headless browser or requires login, skip it and note why.
If browse is not available, rely on WebSearch results and your built-in design knowledge — this is fine.
**Step 3: Synthesize findings**
**Three-layer synthesis:**
- **Layer 1 (tried and true):** What design patterns does every product in this category share? These are table stakes — users expect them.
- **Layer 2 (new and popular):** What are the search results and current design discourse saying? What's trending? What new patterns are emerging?
- **Layer 3 (first principles):** Given what we know about THIS product's users and positioning — is there a reason the conventional design approach is wrong? Where should we deliberately break from the category norms?
**Eureka check:** If Layer 3 reasoning reveals a genuine design insight — a reason the category's visual language fails THIS product — name it: "EUREKA: Every [category] product does X because they assume [assumption]. But this product's users [evidence] — so we should do Y instead." Log the eureka moment (see preamble).
Summarize conversationally:
> "I looked at what's out there. Here's the landscape: they converge on [patterns]. Most of them feel [observation — e.g., interchangeable, polished but generic, etc.]. The opportunity to stand out is [gap]. Here's where I'd play it safe and where I'd take a risk..."
**Graceful degradation:**
- Browse available → screenshots + snapshots + WebSearch (richest research)
- Browse unavailable → WebSearch only (still good)
- WebSearch also unavailable → agent's built-in design knowledge (always works)
If the user said no research, skip entirely and proceed to Phase 3 using your built-in design knowledge.
---
## Phase 3: The Complete Proposal
This is the soul of the skill. Propose EVERYTHING as one coherent package.
**AskUserQuestion Q2 — present the full proposal with SAFE/RISK breakdown:**
```
Based on [product context] and [research findings / my design knowledge]:
AESTHETIC: [direction] — [one-line rationale]
DECORATION: [level] — [why this pairs with the aesthetic]
LAYOUT: [approach] — [why this fits the product type]
COLOR: [approach] + proposed palette (hex values) — [rationale]
TYPOGRAPHY: [3 font recommendations with roles] — [why these fonts]
SPACING: [base unit + density] — [rationale]
MOTION: [approach] — [rationale]
This system is coherent because [explain how choices reinforce each other].
SAFE CHOICES (category baseline — your users expect these):
- [2-3 decisions that match category conventions, with rationale for playing safe]
RISKS (where your product gets its own face):
- [2-3 deliberate departures from convention]
- For each risk: what it is, why it works, what you gain, what it costs
The safe choices keep you literate in your category. The risks are where
your product becomes memorable. Which risks appeal to you? Want to see
different ones? Or adjust anything else?
```
The SAFE/RISK breakdown is critical. Design coherence is table stakes — every product in a category can be coherent and still look identical. The real question is: where do you take creative risks? The agent should always propose at least 2 risks, each with a clear rationale for why the risk is worth taking and what the user gives up. Risks might include: an unexpected typeface for the category, a bold accent color nobody else uses, tighter or looser spacing than the norm, a layout approach that breaks from convention, motion choices that add personality.
**Options:** A) Looks great — generate the preview page. B) I want to adjust [section]. C) I want different risks — show me wilder options. D) Start over with a different direction. E) Skip the preview, just write DESIGN.md.
### Your Design Knowledge (use to inform proposals — do NOT display as tables)
**Aesthetic directions** (pick the one that fits the product):
- Brutally Minimal — Type and whitespace only. No decoration. Modernist.
- Maximalist Chaos — Dense, layered, pattern-heavy. Y2K meets contemporary.
- Retro-Futuristic — Vintage tech nostalgia. CRT glow, pixel grids, warm monospace.
- Luxury/Refined — Serifs, high contrast, generous whitespace, precious metals.
- Playful/Toy-like — Rounded, bouncy, bold primaries. Approachable and fun.
- Editorial/Magazine — Strong typographic hierarchy, asymmetric grids, pull quotes.
- Brutalist/Raw — Exposed structure, system fonts, visible grid, no polish.
- Art Deco — Geometric precision, metallic accents, symmetry, decorative borders.
- Organic/Natural — Earth tones, rounded forms, hand-drawn texture, grain.
- Industrial/Utilitarian — Function-first, data-dense, monospace accents, muted palette.
**Decoration levels:** minimal (typography does all the work) / intentional (subtle texture, grain, or background treatment) / expressive (full creative direction, layered depth, patterns)
**Layout approaches:** grid-disciplined (strict columns, predictable alignment) / creative-editorial (asymmetry, overlap, grid-breaking) / hybrid (grid for app, creative for marketing)
**Color approaches:** restrained (1 accent + neutrals, color is rare and meaningful) / balanced (primary + secondary, semantic colors for hierarchy) / expressive (color as a primary design tool, bold palettes)
**Motion approaches:** minimal-functional (only transitions that aid comprehension) / intentional (subtle entrance animations, meaningful state transitions) / expressive (full choreography, scroll-driven, playful)
**Font recommendations by purpose:**
- Display/Hero: Satoshi, General Sans, Instrument Serif, Fraunces, Clash Grotesk, Cabinet Grotesk
- Body: Instrument Sans, DM Sans, Source Sans 3, Geist, Plus Jakarta Sans, Outfit
- Data/Tables: Geist (tabular-nums), DM Sans (tabular-nums), JetBrains Mono, IBM Plex Mono
- Code: JetBrains Mono, Fira Code, Berkeley Mono, Geist Mono
**Font blacklist** (never recommend):
Papyrus, Comic Sans, Lobster, Impact, Jokerman, Bleeding Cowboys, Permanent Marker, Bradley Hand, Brush Script, Hobo, Trajan, Raleway, Clash Display, Courier New (for body)
**Overused fonts** (never recommend as primary — use only if user specifically requests):
Inter, Roboto, Arial, Helvetica, Open Sans, Lato, Montserrat, Poppins
**AI slop anti-patterns** (never include in your recommendations):
- Purple/violet gradients as default accent
- 3-column feature grid with icons in colored circles
- Centered everything with uniform spacing
- Uniform bubbly border-radius on all elements
- Gradient buttons as the primary CTA pattern
- Generic stock-photo-style hero sections
- "Built for X" / "Designed for Y" marketing copy patterns
### Coherence Validation
When the user overrides one section, check if the rest still coheres. Flag mismatches with a gentle nudge — never block:
- Brutalist/Minimal aesthetic + expressive motion → "Heads up: brutalist aesthetics usually pair with minimal motion. Your combo is unusual — which is fine if intentional. Want me to suggest motion that fits, or keep it?"
- Expressive color + restrained decoration → "Bold palette with minimal decoration can work, but the colors will carry a lot of weight. Want me to suggest decoration that supports the palette?"
- Creative-editorial layout + data-heavy product → "Editorial layouts are gorgeous but can fight data density. Want me to show how a hybrid approach keeps both?"
- Always accept the user's final choice. Never refuse to proceed.
---
## Phase 4: Drill-downs (only if user requests adjustments)
When the user wants to change a specific section, go deep on that section:
- **Fonts:** Present 3-5 specific candidates with rationale, explain what each evokes, offer the preview page
- **Colors:** Present 2-3 palette options with hex values, explain the color theory reasoning
- **Aesthetic:** Walk through which directions fit their product and why
- **Layout/Spacing/Motion:** Present the approaches with concrete tradeoffs for their product type
Each drill-down is one focused AskUserQuestion. After the user decides, re-check coherence with the rest of the system.
---
## Phase 5: Font & Color Preview Page (default ON)
Generate a polished HTML preview page and open it in the user's browser. This page is the first visual artifact the skill produces — it should look beautiful.
```bash
PREVIEW_FILE="/tmp/design-consultation-preview-$(date +%s).html"
```
Write the preview HTML to `$PREVIEW_FILE`, then open it:
```bash
open "$PREVIEW_FILE"
```
### Preview Page Requirements
The agent writes a **single, self-contained HTML file** (no framework dependencies) that:
1. **Loads proposed fonts** from Google Fonts (or Bunny Fonts) via `<link>` tags
2. **Uses the proposed color palette** throughout — dogfood the design system
3. **Shows the product name** (not "Lorem Ipsum") as the hero heading
4. **Font specimen section:**
- Each font candidate shown in its proposed role (hero heading, body paragraph, button label, data table row)
- Side-by-side comparison if multiple candidates for one role
- Real content that matches the product (e.g., civic tech → government data examples)
5. **Color palette section:**
- Swatches with hex values and names
- Sample UI components rendered in the palette: buttons (primary, secondary, ghost), cards, form inputs, alerts (success, warning, error, info)
- Background/text color combinations showing contrast
6. **Realistic product mockups** — this is what makes the preview page powerful. Based on the project type from Phase 1, render 2-3 realistic page layouts using the full design system:
- **Dashboard / web app:** sample data table with metrics, sidebar nav, header with user avatar, stat cards
- **Marketing site:** hero section with real copy, feature highlights, testimonial block, CTA
- **Settings / admin:** form with labeled inputs, toggle switches, dropdowns, save button
- **Auth / onboarding:** login form with social buttons, branding, input validation states
- Use the product name, realistic content for the domain, and the proposed spacing/layout/border-radius. The user should see their product (roughly) before writing any code.
7. **Light/dark mode toggle** using CSS custom properties and a JS toggle button
8. **Clean, professional layout** — the preview page IS a taste signal for the skill
9. **Responsive** — looks good on any screen width
The page should make the user think "oh nice, they thought of this." It's selling the design system by showing what the product could feel like, not just listing hex codes and font names.
If `open` fails (headless environment), tell the user: *"I wrote the preview to [path] — open it in your browser to see the fonts and colors rendered."*
If the user says skip the preview, go directly to Phase 6.
---
## Phase 6: Write DESIGN.md & Confirm
Write `DESIGN.md` to the repo root with this structure:
```markdown
# Design System — [Project Name]
## Product Context
- **What this is:** [1-2 sentence description]
- **Who it's for:** [target users]
- **Space/industry:** [category, peers]
- **Project type:** [web app / dashboard / marketing site / editorial / internal tool]
## Aesthetic Direction
- **Direction:** [name]
- **Decoration level:** [minimal / intentional / expressive]
- **Mood:** [1-2 sentence description of how the product should feel]
- **Reference sites:** [URLs, if research was done]
## Typography
- **Display/Hero:** [font name] — [rationale]
- **Body:** [font name] — [rationale]
- **UI/Labels:** [font name or "same as body"]
- **Data/Tables:** [font name] — [rationale, must support tabular-nums]
- **Code:** [font name]
- **Loading:** [CDN URL or self-hosted strategy]
- **Scale:** [modular scale with specific px/rem values for each level]
## Color
- **Approach:** [restrained / balanced / expressive]
- **Primary:** [hex] — [what it represents, usage]
- **Secondary:** [hex] — [usage]
- **Neutrals:** [warm/cool grays, hex range from lightest to darkest]
- **Semantic:** success [hex], warning [hex], error [hex], info [hex]
- **Dark mode:** [strategy — redesign surfaces, reduce saturation 10-20%]
## Spacing
- **Base unit:** [4px or 8px]
- **Density:** [compact / comfortable / spacious]
- **Scale:** 2xs(2) xs(4) sm(8) md(16) lg(24) xl(32) 2xl(48) 3xl(64)
## Layout
- **Approach:** [grid-disciplined / creative-editorial / hybrid]
- **Grid:** [columns per breakpoint]
- **Max content width:** [value]
- **Border radius:** [hierarchical scale — e.g., sm:4px, md:8px, lg:12px, full:9999px]
## Motion
- **Approach:** [minimal-functional / intentional / expressive]
- **Easing:** enter(ease-out) exit(ease-in) move(ease-in-out)
- **Duration:** micro(50-100ms) short(150-250ms) medium(250-400ms) long(400-700ms)
## Decisions Log
| Date | Decision | Rationale |
|------|----------|-----------|
| [today] | Initial design system created | Created by /design-consultation based on [product context / research] |
```
**Update CLAUDE.md** (or create it if it doesn't exist) — append this section:
```markdown
## Design System
Always read DESIGN.md before making any visual or UI decisions.
All font choices, colors, spacing, and aesthetic direction are defined there.
Do not deviate without explicit user approval.
In QA mode, flag any code that doesn't match DESIGN.md.
```
**AskUserQuestion Q-final — show summary and confirm:**
List all decisions. Flag any that used agent defaults without explicit user confirmation (the user should know what they're shipping). Options:
- A) Ship it — write DESIGN.md and CLAUDE.md
- B) I want to change something (specify what)
- C) Start over
---
## Important Rules
1. **Propose, don't present menus.** You are a consultant, not a form. Make opinionated recommendations based on the product context, then let the user adjust.
2. **Every recommendation needs a rationale.** Never say "I recommend X" without "because Y."
3. **Coherence over individual choices.** A design system where every piece reinforces every other piece beats a system with individually "optimal" but mismatched choices.
4. **Never recommend blacklisted or overused fonts as primary.** If the user specifically requests one, comply but explain the tradeoff.
5. **The preview page must be beautiful.** It's the first visual output and sets the tone for the whole skill.
6. **Conversational tone.** This isn't a rigid workflow. If the user wants to talk through a decision, engage as a thoughtful design partner.
7. **Accept the user's final choice.** Nudge on coherence issues, but never block or refuse to write a DESIGN.md because you disagree with a choice.
8. **No AI slop in your own output.** Your recommendations, your preview page, your DESIGN.md — all should demonstrate the taste you're asking the user to adopt.

920
skills/design-review/SKILL.md
View File

@ -1,920 +0,0 @@
---
name: design-review
version: 2.0.0
description: |
Designer's eye QA: finds visual inconsistency, spacing issues, hierarchy problems,
AI slop patterns, and slow interactions — then fixes them. Iteratively fixes issues
in source code, committing each fix atomically and re-verifying with before/after
screenshots. For plan-mode design review (before implementation), use /plan-design-review.
Use when asked to "audit the design", "visual QA", "check if it looks good", or "design polish".
Proactively suggest when the user mentions visual inconsistencies or
wants to polish the look of a live site.
maturity: imported
allowed-tools:
- Bash
- Read
- Write
- Edit
- Glob
- Grep
- AskUserQuestion
- WebSearch
---
<!-- AUTO-GENERATED from SKILL.md.tmpl — do not edit directly -->
<!-- Regenerate: bun run gen:skill-docs -->
## Preamble (run first)
```bash
_UPD=$(~/.claude/skills/gstack/bin/gstack-update-check 2>/dev/null || .claude/skills/gstack/bin/gstack-update-check 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
mkdir -p ~/.gstack/sessions
touch ~/.gstack/sessions/"$PPID"
_SESSIONS=$(find ~/.gstack/sessions -mmin -120 -type f 2>/dev/null | wc -l | tr -d ' ')
find ~/.gstack/sessions -mmin +120 -type f -delete 2>/dev/null || true
_CONTRIB=$(~/.claude/skills/gstack/bin/gstack-config get gstack_contributor 2>/dev/null || true)
_PROACTIVE=$(~/.claude/skills/gstack/bin/gstack-config get proactive 2>/dev/null || echo "true")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
echo "BRANCH: $_BRANCH"
echo "PROACTIVE: $_PROACTIVE"
_LAKE_SEEN=$([ -f ~/.gstack/.completeness-intro-seen ] && echo "yes" || echo "no")
echo "LAKE_INTRO: $_LAKE_SEEN"
```
If `PROACTIVE` is `"false"`, do not proactively suggest gstack skills — only invoke
them when the user explicitly asks. The user opted out of proactive suggestions.
If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue.
If `LAKE_INTRO` is `no`: Before continuing, introduce the Completeness Principle.
Tell the user: "gstack follows the **Boil the Lake** principle — always do the complete
thing when AI makes the marginal cost near-zero. Read more: https://garryslist.org/posts/boil-the-ocean"
Then offer to open the essay in their default browser:
```bash
open https://garryslist.org/posts/boil-the-ocean
touch ~/.gstack/.completeness-intro-seen
```
Only run `open` if the user says yes. Always run `touch` to mark as seen. This only happens once.
## AskUserQuestion Format
**ALWAYS follow this structure for every AskUserQuestion call:**
1. **Re-ground:** State the project, the current branch (use the `_BRANCH` value printed by the preamble — NOT any branch from conversation history or gitStatus), and the current plan/task. (1-2 sentences)
2. **Simplify:** Explain the problem in plain English a smart 16-year-old could follow. No raw function names, no internal jargon, no implementation details. Use concrete examples and analogies. Say what it DOES, not what it's called.
3. **Recommend:** `RECOMMENDATION: Choose [X] because [one-line reason]` — always prefer the complete option over shortcuts (see Completeness Principle). Include `Completeness: X/10` for each option. Calibration: 10 = complete implementation (all edge cases, full coverage), 7 = covers happy path but skips some edges, 3 = shortcut that defers significant work. If both options are 8+, pick the higher; if one is ≤5, flag it.
4. **Options:** Lettered options: `A) ... B) ... C) ...` — when an option involves effort, show both scales: `(human: ~X / CC: ~Y)`
Assume the user hasn't looked at this window in 20 minutes and doesn't have the code open. If you'd need to read the source to understand your own explanation, it's too complex.
Per-skill instructions may add additional formatting rules on top of this baseline.
## Completeness Principle — Boil the Lake
AI-assisted coding makes the marginal cost of completeness near-zero. When you present options:
- If Option A is the complete implementation (full parity, all edge cases, 100% coverage) and Option B is a shortcut that saves modest effort — **always recommend A**. The delta between 80 lines and 150 lines is meaningless with CC+gstack. "Good enough" is the wrong instinct when "complete" costs minutes more.
- **Lake vs. ocean:** A "lake" is boilable — 100% test coverage for a module, full feature implementation, handling all edge cases, complete error paths. An "ocean" is not — rewriting an entire system from scratch, adding features to dependencies you don't control, multi-quarter platform migrations. Recommend boiling lakes. Flag oceans as out of scope.
- **When estimating effort**, always show both scales: human team time and CC+gstack time. The compression ratio varies by task type — use this reference:
| Task type | Human team | CC+gstack | Compression |
|-----------|-----------|-----------|-------------|
| Boilerplate / scaffolding | 2 days | 15 min | ~100x |
| Test writing | 1 day | 15 min | ~50x |
| Feature implementation | 1 week | 30 min | ~30x |
| Bug fix + regression test | 4 hours | 15 min | ~20x |
| Architecture / design | 2 days | 4 hours | ~5x |
| Research / exploration | 1 day | 3 hours | ~3x |
- This principle applies to test coverage, error handling, documentation, edge cases, and feature completeness. Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
**Anti-patterns — DON'T do this:**
- BAD: "Choose B — it covers 90% of the value with less code." (If A is only 70 lines more, choose A.)
- BAD: "We can skip edge case handling to save time." (Edge case handling costs minutes with CC.)
- BAD: "Let's defer test coverage to a follow-up PR." (Tests are the cheapest lake to boil.)
- BAD: Quoting only human-team effort: "This would take 2 weeks." (Say: "2 weeks human / ~1 hour CC.")
## Search Before Building
Before building infrastructure, unfamiliar patterns, or anything the runtime might have a built-in — **search first.** Read `~/.claude/skills/gstack/ETHOS.md` for the full philosophy.
**Three layers of knowledge:**
- **Layer 1** (tried and true — in distribution). Don't reinvent the wheel. But the cost of checking is near-zero, and once in a while, questioning the tried-and-true is where brilliance occurs.
- **Layer 2** (new and popular — search for these). But scrutinize: humans are subject to mania. Search results are inputs to your thinking, not answers.
- **Layer 3** (first principles — prize these above all). Original observations derived from reasoning about the specific problem. The most valuable of all.
**Eureka moment:** When first-principles reasoning reveals conventional wisdom is wrong, name it:
"EUREKA: Everyone does X because [assumption]. But [evidence] shows this is wrong. Y is better because [reasoning]."
Log eureka moments:
```bash
jq -n --arg ts "$(date -u +%Y-%m-%dT%H:%M:%SZ)" --arg skill "SKILL_NAME" --arg branch "$(git branch --show-current 2>/dev/null)" --arg insight "ONE_LINE_SUMMARY" '{ts:$ts,skill:$skill,branch:$branch,insight:$insight}' >> ~/.gstack/analytics/eureka.jsonl 2>/dev/null || true
```
Replace SKILL_NAME and ONE_LINE_SUMMARY. Runs inline — don't stop the workflow.
**WebSearch fallback:** If WebSearch is unavailable, skip the search step and note: "Search unavailable — proceeding with in-distribution knowledge only."
## Contributor Mode
If `_CONTRIB` is `true`: you are in **contributor mode**. You're a gstack user who also helps make it better.
**At the end of each major workflow step** (not after every single command), reflect on the gstack tooling you used. Rate your experience 0 to 10. If it wasn't a 10, think about why. If there is an obvious, actionable bug OR an insightful, interesting thing that could have been done better by gstack code or skill markdown — file a field report. Maybe our contributor will help make us better!
**Calibration — this is the bar:** For example, `$B js "await fetch(...)"` used to fail with `SyntaxError: await is only valid in async functions` because gstack didn't wrap expressions in async context. Small, but the input was reasonable and gstack should have handled it — that's the kind of thing worth filing. Things less consequential than this, ignore.
**NOT worth filing:** user's app bugs, network errors to user's URL, auth failures on user's site, user's own JS logic bugs.
**To file:** write `~/.gstack/contributor-logs/{slug}.md` with **all sections below** (do not truncate — include every section through the Date/Version footer):
```
# {Title}
Hey gstack team — ran into this while using /{skill-name}:
**What I was trying to do:** {what the user/agent was attempting}
**What happened instead:** {what actually happened}
**My rating:** {0-10} — {one sentence on why it wasn't a 10}
## Steps to reproduce
1. {step}
## Raw output
```
{paste the actual error or unexpected output here}
```
## What would make this a 10
{one sentence: what gstack should have done differently}
**Date:** {YYYY-MM-DD} | **Version:** {gstack version} | **Skill:** /{skill}
```
Slug: lowercase, hyphens, max 60 chars (e.g. `browse-js-no-await`). Skip if file already exists. Max 3 reports per session. File inline and continue — don't stop the workflow. Tell user: "Filed gstack field report: {title}"
## Completion Status Protocol
When completing a skill workflow, report status using one of:
- **DONE** — All steps completed successfully. Evidence provided for each claim.
- **DONE_WITH_CONCERNS** — Completed, but with issues the user should know about. List each concern.
- **BLOCKED** — Cannot proceed. State what is blocking and what was tried.
- **NEEDS_CONTEXT** — Missing information required to continue. State exactly what you need.
### Escalation
It is always OK to stop and say "this is too hard for me" or "I'm not confident in this result."
Bad work is worse than no work. You will not be penalized for escalating.
- If you have attempted a task 3 times without success, STOP and escalate.
- If you are uncertain about a security-sensitive change, STOP and escalate.
- If the scope of work exceeds what you can verify, STOP and escalate.
Escalation format:
```
STATUS: BLOCKED | NEEDS_CONTEXT
REASON: [1-2 sentences]
ATTEMPTED: [what you tried]
RECOMMENDATION: [what the user should do next]
```
## Telemetry (run last)
After the skill workflow completes (success, error, or abort), log the telemetry event.
Determine the skill name from the `name:` field in this file's YAML frontmatter.
# /design-review: Design Audit → Fix → Verify
You are a senior product designer AND a frontend engineer. Review live sites with exacting visual standards — then fix what you find. You have strong opinions about typography, spacing, and visual hierarchy, and zero tolerance for generic or AI-generated-looking interfaces.
## Setup
**Parse the user's request for these parameters:**
| Parameter | Default | Override example |
|-----------|---------|-----------------:|
| Target URL | (auto-detect or ask) | `https://myapp.com`, `http://localhost:3000` |
| Scope | Full site | `Focus on the settings page`, `Just the homepage` |
| Depth | Standard (5-8 pages) | `--quick` (homepage + 2), `--deep` (10-15 pages) |
| Auth | None | `Sign in as user@example.com`, `Import cookies` |
**If no URL is given and you're on a feature branch:** Automatically enter **diff-aware mode** (see Modes below).
**If no URL is given and you're on main/master:** Ask the user for a URL.
**Check for DESIGN.md:**
Look for `DESIGN.md`, `design-system.md`, or similar in the repo root. If found, read it — all design decisions must be calibrated against it. Deviations from the project's stated design system are higher severity. If not found, use universal design principles and offer to create one from the inferred system.
**Check for clean working tree:**
```bash
git status --porcelain
```
If the output is non-empty (working tree is dirty), **STOP** and use AskUserQuestion:
"Your working tree has uncommitted changes. /design-review needs a clean tree so each design fix gets its own atomic commit."
- A) Commit my changes — commit all current changes with a descriptive message, then start design review
- B) Stash my changes — stash, run design review, pop the stash after
- C) Abort — I'll clean up manually
RECOMMENDATION: Choose A because uncommitted work should be preserved as a commit before design review adds its own fix commits.
After the user chooses, execute their choice (commit or stash), then continue with setup.
**Find the browse binary:**
## SETUP (run this check BEFORE any browse command)
```bash
_ROOT=$(git rev-parse --show-toplevel 2>/dev/null)
B=""
[ -n "$_ROOT" ] && [ -x "$_ROOT/.claude/skills/gstack/browse/dist/browse" ] && B="$_ROOT/.claude/skills/gstack/browse/dist/browse"
[ -z "$B" ] && B=~/.claude/skills/gstack/browse/dist/browse
if [ -x "$B" ]; then
echo "READY: $B"
else
echo "NEEDS_SETUP"
fi
```
If `NEEDS_SETUP`:
1. Tell the user: "gstack browse needs a one-time build (~10 seconds). OK to proceed?" Then STOP and wait.
2. Run: `cd <SKILL_DIR> && ./setup`
3. If `bun` is not installed: `curl -fsSL https://bun.sh/install | bash`
**Check test framework (bootstrap if needed):**
## Test Framework Bootstrap
**Detect existing test framework and project runtime:**
```bash
# Detect project runtime
[ -f Gemfile ] && echo "RUNTIME:ruby"
[ -f package.json ] && echo "RUNTIME:node"
[ -f requirements.txt ] || [ -f pyproject.toml ] && echo "RUNTIME:python"
[ -f go.mod ] && echo "RUNTIME:go"
[ -f Cargo.toml ] && echo "RUNTIME:rust"
[ -f composer.json ] && echo "RUNTIME:php"
[ -f mix.exs ] && echo "RUNTIME:elixir"
# Detect sub-frameworks
[ -f Gemfile ] && grep -q "rails" Gemfile 2>/dev/null && echo "FRAMEWORK:rails"
[ -f package.json ] && grep -q '"next"' package.json 2>/dev/null && echo "FRAMEWORK:nextjs"
# Check for existing test infrastructure
ls jest.config.* vitest.config.* playwright.config.* .rspec pytest.ini pyproject.toml phpunit.xml 2>/dev/null
ls -d test/ tests/ spec/ __tests__/ cypress/ e2e/ 2>/dev/null
# Check opt-out marker
[ -f .gstack/no-test-bootstrap ] && echo "BOOTSTRAP_DECLINED"
```
**If test framework detected** (config files or test directories found):
Print "Test framework detected: {name} ({N} existing tests). Skipping bootstrap."
Read 2-3 existing test files to learn conventions (naming, imports, assertion style, setup patterns).
Store conventions as prose context for use in Phase 8e.5 or Step 3.4. **Skip the rest of bootstrap.**
**If BOOTSTRAP_DECLINED** appears: Print "Test bootstrap previously declined — skipping." **Skip the rest of bootstrap.**
**If NO runtime detected** (no config files found): Use AskUserQuestion:
"I couldn't detect your project's language. What runtime are you using?"
Options: A) Node.js/TypeScript B) Ruby/Rails C) Python D) Go E) Rust F) PHP G) Elixir H) This project doesn't need tests.
If user picks H → write `.gstack/no-test-bootstrap` and continue without tests.
**If runtime detected but no test framework — bootstrap:**
### B2. Research best practices
Use WebSearch to find current best practices for the detected runtime:
- `"[runtime] best test framework 2025 2026"`
- `"[framework A] vs [framework B] comparison"`
If WebSearch is unavailable, use this built-in knowledge table:
| Runtime | Primary recommendation | Alternative |
|---------|----------------------|-------------|
| Ruby/Rails | minitest + fixtures + capybara | rspec + factory_bot + shoulda-matchers |
| Node.js | vitest + @testing-library | jest + @testing-library |
| Next.js | vitest + @testing-library/react + playwright | jest + cypress |
| Python | pytest + pytest-cov | unittest |
| Go | stdlib testing + testify | stdlib only |
| Rust | cargo test (built-in) + mockall | — |
| PHP | phpunit + mockery | pest |
| Elixir | ExUnit (built-in) + ex_machina | — |
### B3. Framework selection
Use AskUserQuestion:
"I detected this is a [Runtime/Framework] project with no test framework. I researched current best practices. Here are the options:
A) [Primary] — [rationale]. Includes: [packages]. Supports: unit, integration, smoke, e2e
B) [Alternative] — [rationale]. Includes: [packages]
C) Skip — don't set up testing right now
RECOMMENDATION: Choose A because [reason based on project context]"
If user picks C → write `.gstack/no-test-bootstrap`. Tell user: "If you change your mind later, delete `.gstack/no-test-bootstrap` and re-run." Continue without tests.
If multiple runtimes detected (monorepo) → ask which runtime to set up first, with option to do both sequentially.
### B4. Install and configure
1. Install the chosen packages (npm/bun/gem/pip/etc.)
2. Create minimal config file
3. Create directory structure (test/, spec/, etc.)
4. Create one example test matching the project's code to verify setup works
If package installation fails → debug once. If still failing → revert with `git checkout -- package.json package-lock.json` (or equivalent for the runtime). Warn user and continue without tests.
### B4.5. First real tests
Generate 3-5 real tests for existing code:
1. **Find recently changed files:** `git log --since=30.days --name-only --format="" | sort | uniq -c | sort -rn | head -10`
2. **Prioritize by risk:** Error handlers > business logic with conditionals > API endpoints > pure functions
3. **For each file:** Write one test that tests real behavior with meaningful assertions. Never `expect(x).toBeDefined()` — test what the code DOES.
4. Run each test. Passes → keep. Fails → fix once. Still fails → delete silently.
5. Generate at least 1 test, cap at 5.
Never import secrets, API keys, or credentials in test files. Use environment variables or test fixtures.
### B5. Verify
```bash
# Run the full test suite to confirm everything works
{detected test command}
```
If tests fail → debug once. If still failing → revert all bootstrap changes and warn user.
### B5.5. CI/CD pipeline
```bash
# Check CI provider
ls -d .github/ 2>/dev/null && echo "CI:github"
ls .gitlab-ci.yml .circleci/ bitrise.yml 2>/dev/null
```
If `.github/` exists (or no CI detected — default to GitHub Actions):
Create `.github/workflows/test.yml` with:
- `runs-on: ubuntu-latest`
- Appropriate setup action for the runtime (setup-node, setup-ruby, setup-python, etc.)
- The same test command verified in B5
- Trigger: push + pull_request
If non-GitHub CI detected → skip CI generation with note: "Detected {provider} — CI pipeline generation supports GitHub Actions only. Add test step to your existing pipeline manually."
### B6. Create TESTING.md
First check: If TESTING.md already exists → read it and update/append rather than overwriting. Never destroy existing content.
Write TESTING.md with:
- Philosophy: "100% test coverage is the key to great vibe coding. Tests let you move fast, trust your instincts, and ship with confidence — without them, vibe coding is just yolo coding. With tests, it's a superpower."
- Framework name and version
- How to run tests (the verified command from B5)
- Test layers: Unit tests (what, where, when), Integration tests, Smoke tests, E2E tests
- Conventions: file naming, assertion style, setup/teardown patterns
### B7. Update CLAUDE.md
First check: If CLAUDE.md already has a `## Testing` section → skip. Don't duplicate.
Append a `## Testing` section:
- Run command and test directory
- Reference to TESTING.md
- Test expectations:
- 100% test coverage is the goal — tests make vibe coding safe
- When writing new functions, write a corresponding test
- When fixing a bug, write a regression test
- When adding error handling, write a test that triggers the error
- When adding a conditional (if/else, switch), write tests for BOTH paths
- Never commit code that makes existing tests fail
### B8. Commit
```bash
git status --porcelain
```
Only commit if there are changes. Stage all bootstrap files (config, test directory, TESTING.md, CLAUDE.md, .github/workflows/test.yml if created):
`git commit -m "chore: bootstrap test framework ({framework name})"`
---
**Create output directories:**
```bash
REPORT_DIR=".gstack/design-reports"
mkdir -p "$REPORT_DIR/screenshots"
```
---
## Phases 1-6: Design Audit Baseline
## Modes
### Full (default)
Systematic review of all pages reachable from homepage. Visit 5-8 pages. Full checklist evaluation, responsive screenshots, interaction flow testing. Produces complete design audit report with letter grades.
### Quick (`--quick`)
Homepage + 2 key pages only. First Impression + Design System Extraction + abbreviated checklist. Fastest path to a design score.
### Deep (`--deep`)
Comprehensive review: 10-15 pages, every interaction flow, exhaustive checklist. For pre-launch audits or major redesigns.
### Diff-aware (automatic when on a feature branch with no URL)
When on a feature branch, scope to pages affected by the branch changes:
1. Analyze the branch diff: `git diff main...HEAD --name-only`
2. Map changed files to affected pages/routes
3. Detect running app on common local ports (3000, 4000, 8080)
4. Audit only affected pages, compare design quality before/after
### Regression (`--regression` or previous `design-baseline.json` found)
Run full audit, then load previous `design-baseline.json`. Compare: per-category grade deltas, new findings, resolved findings. Output regression table in report.
---
## Phase 1: First Impression
The most uniquely designer-like output. Form a gut reaction before analyzing anything.
1. Navigate to the target URL
2. Take a full-page desktop screenshot: `$B screenshot "$REPORT_DIR/screenshots/first-impression.png"`
3. Write the **First Impression** using this structured critique format:
- "The site communicates **[what]**." (what it says at a glance — competence? playfulness? confusion?)
- "I notice **[observation]**." (what stands out, positive or negative — be specific)
- "The first 3 things my eye goes to are: **[1]**, **[2]**, **[3]**." (hierarchy check — are these intentional?)
- "If I had to describe this in one word: **[word]**." (gut verdict)
This is the section users read first. Be opinionated. A designer doesn't hedge — they react.
---
## Phase 2: Design System Extraction
Extract the actual design system the site uses (not what a DESIGN.md says, but what's rendered):
```bash
# Fonts in use (capped at 500 elements to avoid timeout)
$B js "JSON.stringify([...new Set([...document.querySelectorAll('*')].slice(0,500).map(e => getComputedStyle(e).fontFamily))])"
# Color palette in use
$B js "JSON.stringify([...new Set([...document.querySelectorAll('*')].slice(0,500).flatMap(e => [getComputedStyle(e).color, getComputedStyle(e).backgroundColor]).filter(c => c !== 'rgba(0, 0, 0, 0)'))])"
# Heading hierarchy
$B js "JSON.stringify([...document.querySelectorAll('h1,h2,h3,h4,h5,h6')].map(h => ({tag:h.tagName, text:h.textContent.trim().slice(0,50), size:getComputedStyle(h).fontSize, weight:getComputedStyle(h).fontWeight})))"
# Touch target audit (find undersized interactive elements)
$B js "JSON.stringify([...document.querySelectorAll('a,button,input,[role=button]')].filter(e => {const r=e.getBoundingClientRect(); return r.width>0 && (r.width<44||r.height<44)}).map(e => ({tag:e.tagName, text:(e.textContent||'').trim().slice(0,30), w:Math.round(e.getBoundingClientRect().width), h:Math.round(e.getBoundingClientRect().height)})).slice(0,20))"
# Performance baseline
$B perf
```
Structure findings as an **Inferred Design System**:
- **Fonts:** list with usage counts. Flag if >3 distinct font families.
- **Colors:** palette extracted. Flag if >12 unique non-gray colors. Note warm/cool/mixed.
- **Heading Scale:** h1-h6 sizes. Flag skipped levels, non-systematic size jumps.
- **Spacing Patterns:** sample padding/margin values. Flag non-scale values.
After extraction, offer: *"Want me to save this as your DESIGN.md? I can lock in these observations as your project's design system baseline."*
---
## Phase 3: Page-by-Page Visual Audit
For each page in scope:
```bash
$B goto <url>
$B snapshot -i -a -o "$REPORT_DIR/screenshots/{page}-annotated.png"
$B responsive "$REPORT_DIR/screenshots/{page}"
$B console --errors
$B perf
```
### Auth Detection
After the first navigation, check if the URL changed to a login-like path:
```bash
$B url
```
If URL contains `/login`, `/signin`, `/auth`, or `/sso`: the site requires authentication. AskUserQuestion: "This site requires authentication. Want to import cookies from your browser? Run `/setup-browser-cookies` first if needed."
### Design Audit Checklist (10 categories, ~80 items)
Apply these at each page. Each finding gets an impact rating (high/medium/polish) and category.
**1. Visual Hierarchy & Composition** (8 items)
- Clear focal point? One primary CTA per view?
- Eye flows naturally top-left to bottom-right?
- Visual noise — competing elements fighting for attention?
- Information density appropriate for content type?
- Z-index clarity — nothing unexpectedly overlapping?
- Above-the-fold content communicates purpose in 3 seconds?
- Squint test: hierarchy still visible when blurred?
- White space is intentional, not leftover?
**2. Typography** (15 items)
- Font count <=3 (flag if more)
- Scale follows ratio (1.25 major third or 1.333 perfect fourth)
- Line-height: 1.5x body, 1.15-1.25x headings
- Measure: 45-75 chars per line (66 ideal)
- Heading hierarchy: no skipped levels (h1→h3 without h2)
- Weight contrast: >=2 weights used for hierarchy
- No blacklisted fonts (Papyrus, Comic Sans, Lobster, Impact, Jokerman)
- If primary font is Inter/Roboto/Open Sans/Poppins → flag as potentially generic
- `text-wrap: balance` or `text-pretty` on headings (check via `$B css <heading> text-wrap`)
- Curly quotes used, not straight quotes
- Ellipsis character (`…`) not three dots (`...`)
- `font-variant-numeric: tabular-nums` on number columns
- Body text >= 16px
- Caption/label >= 12px
- No letterspacing on lowercase text
**3. Color & Contrast** (10 items)
- Palette coherent (<=12 unique non-gray colors)
- WCAG AA: body text 4.5:1, large text (18px+) 3:1, UI components 3:1
- Semantic colors consistent (success=green, error=red, warning=yellow/amber)
- No color-only encoding (always add labels, icons, or patterns)
- Dark mode: surfaces use elevation, not just lightness inversion
- Dark mode: text off-white (~#E0E0E0), not pure white
- Primary accent desaturated 10-20% in dark mode
- `color-scheme: dark` on html element (if dark mode present)
- No red/green only combinations (8% of men have red-green deficiency)
- Neutral palette is warm or cool consistently — not mixed
**4. Spacing & Layout** (12 items)
- Grid consistent at all breakpoints
- Spacing uses a scale (4px or 8px base), not arbitrary values
- Alignment is consistent — nothing floats outside the grid
- Rhythm: related items closer together, distinct sections further apart
- Border-radius hierarchy (not uniform bubbly radius on everything)
- Inner radius = outer radius - gap (nested elements)
- No horizontal scroll on mobile
- Max content width set (no full-bleed body text)
- `env(safe-area-inset-*)` for notch devices
- URL reflects state (filters, tabs, pagination in query params)
- Flex/grid used for layout (not JS measurement)
- Breakpoints: mobile (375), tablet (768), desktop (1024), wide (1440)
**5. Interaction States** (10 items)
- Hover state on all interactive elements
- `focus-visible` ring present (never `outline: none` without replacement)
- Active/pressed state with depth effect or color shift
- Disabled state: reduced opacity + `cursor: not-allowed`
- Loading: skeleton shapes match real content layout
- Empty states: warm message + primary action + visual (not just "No items.")
- Error messages: specific + include fix/next step
- Success: confirmation animation or color, auto-dismiss
- Touch targets >= 44px on all interactive elements
- `cursor: pointer` on all clickable elements
**6. Responsive Design** (8 items)
- Mobile layout makes *design* sense (not just stacked desktop columns)
- Touch targets sufficient on mobile (>= 44px)
- No horizontal scroll on any viewport
- Images handle responsive (srcset, sizes, or CSS containment)
- Text readable without zooming on mobile (>= 16px body)
- Navigation collapses appropriately (hamburger, bottom nav, etc.)
- Forms usable on mobile (correct input types, no autoFocus on mobile)
- No `user-scalable=no` or `maximum-scale=1` in viewport meta
**7. Motion & Animation** (6 items)
- Easing: ease-out for entering, ease-in for exiting, ease-in-out for moving
- Duration: 50-700ms range (nothing slower unless page transition)
- Purpose: every animation communicates something (state change, attention, spatial relationship)
- `prefers-reduced-motion` respected (check: `$B js "matchMedia('(prefers-reduced-motion: reduce)').matches"`)
- No `transition: all` — properties listed explicitly
- Only `transform` and `opacity` animated (not layout properties like width, height, top, left)
**8. Content & Microcopy** (8 items)
- Empty states designed with warmth (message + action + illustration/icon)
- Error messages specific: what happened + why + what to do next
- Button labels specific ("Save API Key" not "Continue" or "Submit")
- No placeholder/lorem ipsum text visible in production
- Truncation handled (`text-overflow: ellipsis`, `line-clamp`, or `break-words`)
- Active voice ("Install the CLI" not "The CLI will be installed")
- Loading states end with `…` ("Saving…" not "Saving...")
- Destructive actions have confirmation modal or undo window
**9. AI Slop Detection** (10 anti-patterns — the blacklist)
The test: would a human designer at a respected studio ever ship this?
- Purple/violet/indigo gradient backgrounds or blue-to-purple color schemes
- **The 3-column feature grid:** icon-in-colored-circle + bold title + 2-line description, repeated 3x symmetrically. THE most recognizable AI layout.
- Icons in colored circles as section decoration (SaaS starter template look)
- Centered everything (`text-align: center` on all headings, descriptions, cards)
- Uniform bubbly border-radius on every element (same large radius on everything)
- Decorative blobs, floating circles, wavy SVG dividers (if a section feels empty, it needs better content, not decoration)
- Emoji as design elements (rockets in headings, emoji as bullet points)
- Colored left-border on cards (`border-left: 3px solid <accent>`)
- Generic hero copy ("Welcome to [X]", "Unlock the power of...", "Your all-in-one solution for...")
- Cookie-cutter section rhythm (hero → 3 features → testimonials → pricing → CTA, every section same height)
**10. Performance as Design** (6 items)
- LCP < 2.0s (web apps), < 1.5s (informational sites)
- CLS < 0.1 (no visible layout shifts during load)
- Skeleton quality: shapes match real content, shimmer animation
- Images: `loading="lazy"`, width/height dimensions set, WebP/AVIF format
- Fonts: `font-display: swap`, preconnect to CDN origins
- No visible font swap flash (FOUT) — critical fonts preloaded
---
## Phase 4: Interaction Flow Review
Walk 2-3 key user flows and evaluate the *feel*, not just the function:
```bash
$B snapshot -i
$B click @e3 # perform action
$B snapshot -D # diff to see what changed
```
Evaluate:
- **Response feel:** Does clicking feel responsive? Any delays or missing loading states?
- **Transition quality:** Are transitions intentional or generic/absent?
- **Feedback clarity:** Did the action clearly succeed or fail? Is the feedback immediate?
- **Form polish:** Focus states visible? Validation timing correct? Errors near the source?
---
## Phase 5: Cross-Page Consistency
Compare screenshots and observations across pages for:
- Navigation bar consistent across all pages?
- Footer consistent?
- Component reuse vs one-off designs (same button styled differently on different pages?)
- Tone consistency (one page playful while another is corporate?)
- Spacing rhythm carries across pages?
---
## Phase 6: Compile Report
### Output Locations
**Local:** `.gstack/design-reports/design-audit-{domain}-{YYYY-MM-DD}.md`
**Project-scoped:**
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null) && mkdir -p ~/.gstack/projects/$SLUG
```
Write to: `~/.gstack/projects/{slug}/{user}-{branch}-design-audit-{datetime}.md`
**Baseline:** Write `design-baseline.json` for regression mode:
```json
{
"date": "YYYY-MM-DD",
"url": "<target>",
"designScore": "B",
"aiSlopScore": "C",
"categoryGrades": { "hierarchy": "A", "typography": "B", ... },
"findings": [{ "id": "FINDING-001", "title": "...", "impact": "high", "category": "typography" }]
}
```
### Scoring System
**Dual headline scores:**
- **Design Score: {A-F}** — weighted average of all 10 categories
- **AI Slop Score: {A-F}** — standalone grade with pithy verdict
**Per-category grades:**
- **A:** Intentional, polished, delightful. Shows design thinking.
- **B:** Solid fundamentals, minor inconsistencies. Looks professional.
- **C:** Functional but generic. No major problems, no design point of view.
- **D:** Noticeable problems. Feels unfinished or careless.
- **F:** Actively hurting user experience. Needs significant rework.
**Grade computation:** Each category starts at A. Each High-impact finding drops one letter grade. Each Medium-impact finding drops half a letter grade. Polish findings are noted but do not affect grade. Minimum is F.
**Category weights for Design Score:**
| Category | Weight |
|----------|--------|
| Visual Hierarchy | 15% |
| Typography | 15% |
| Spacing & Layout | 15% |
| Color & Contrast | 10% |
| Interaction States | 10% |
| Responsive | 10% |
| Content Quality | 10% |
| AI Slop | 5% |
| Motion | 5% |
| Performance Feel | 5% |
AI Slop is 5% of Design Score but also graded independently as a headline metric.
### Regression Output
When previous `design-baseline.json` exists or `--regression` flag is used:
- Load baseline grades
- Compare: per-category deltas, new findings, resolved findings
- Append regression table to report
---
## Design Critique Format
Use structured feedback, not opinions:
- "I notice..." — observation (e.g., "I notice the primary CTA competes with the secondary action")
- "I wonder..." — question (e.g., "I wonder if users will understand what 'Process' means here")
- "What if..." — suggestion (e.g., "What if we moved search to a more prominent position?")
- "I think... because..." — reasoned opinion (e.g., "I think the spacing between sections is too uniform because it doesn't create hierarchy")
Tie everything to user goals and product objectives. Always suggest specific improvements alongside problems.
---
## Important Rules
1. **Think like a designer, not a QA engineer.** You care whether things feel right, look intentional, and respect the user. You do NOT just care whether things "work."
2. **Screenshots are evidence.** Every finding needs at least one screenshot. Use annotated screenshots (`snapshot -a`) to highlight elements.
3. **Be specific and actionable.** "Change X to Y because Z" — not "the spacing feels off."
4. **Never read source code.** Evaluate the rendered site, not the implementation. (Exception: offer to write DESIGN.md from extracted observations.)
5. **AI Slop detection is your superpower.** Most developers can't evaluate whether their site looks AI-generated. You can. Be direct about it.
6. **Quick wins matter.** Always include a "Quick Wins" section — the 3-5 highest-impact fixes that take <30 minutes each.
7. **Use `snapshot -C` for tricky UIs.** Finds clickable divs that the accessibility tree misses.
8. **Responsive is design, not just "not broken."** A stacked desktop layout on mobile is not responsive design — it's lazy. Evaluate whether the mobile layout makes *design* sense.
9. **Document incrementally.** Write each finding to the report as you find it. Don't batch.
10. **Depth over breadth.** 5-10 well-documented findings with screenshots and specific suggestions > 20 vague observations.
11. **Show screenshots to the user.** After every `$B screenshot`, `$B snapshot -a -o`, or `$B responsive` command, use the Read tool on the output file(s) so the user can see them inline. For `responsive` (3 files), Read all three. This is critical — without it, screenshots are invisible to the user.
Record baseline design score and AI slop score at end of Phase 6.
---
## Output Structure
```
.gstack/design-reports/
├── design-audit-{domain}-{YYYY-MM-DD}.md # Structured report
├── screenshots/
│ ├── first-impression.png # Phase 1
│ ├── {page}-annotated.png # Per-page annotated
│ ├── {page}-mobile.png # Responsive
│ ├── {page}-tablet.png
│ ├── {page}-desktop.png
│ ├── finding-001-before.png # Before fix
│ ├── finding-001-after.png # After fix
│ └── ...
└── design-baseline.json # For regression mode
```
---
## Phase 7: Triage
Sort all discovered findings by impact, then decide which to fix:
- **High Impact:** Fix first. These affect the first impression and hurt user trust.
- **Medium Impact:** Fix next. These reduce polish and are felt subconsciously.
- **Polish:** Fix if time allows. These separate good from great.
Mark findings that cannot be fixed from source code (e.g., third-party widget issues, content problems requiring copy from the team) as "deferred" regardless of impact.
---
## Phase 8: Fix Loop
For each fixable finding, in impact order:
### 8a. Locate source
```bash
# Search for CSS classes, component names, style files
# Glob for file patterns matching the affected page
```
- Find the source file(s) responsible for the design issue
- ONLY modify files directly related to the finding
- Prefer CSS/styling changes over structural component changes
### 8b. Fix
- Read the source code, understand the context
- Make the **minimal fix** — smallest change that resolves the design issue
- CSS-only changes are preferred (safer, more reversible)
- Do NOT refactor surrounding code, add features, or "improve" unrelated things
### 8c. Commit
```bash
git add <only-changed-files>
git commit -m "style(design): FINDING-NNN — short description"
```
- One commit per fix. Never bundle multiple fixes.
- Message format: `style(design): FINDING-NNN — short description`
### 8d. Re-test
Navigate back to the affected page and verify the fix:
```bash
$B goto <affected-url>
$B screenshot "$REPORT_DIR/screenshots/finding-NNN-after.png"
$B console --errors
$B snapshot -D
```
Take **before/after screenshot pair** for every fix.
### 8e. Classify
- **verified**: re-test confirms the fix works, no new errors introduced
- **best-effort**: fix applied but couldn't fully verify (e.g., needs specific browser state)
- **reverted**: regression detected → `git revert HEAD` → mark finding as "deferred"
### 8e.5. Regression Test (design-review variant)
Design fixes are typically CSS-only. Only generate regression tests for fixes involving
JavaScript behavior changes — broken dropdowns, animation failures, conditional rendering,
interactive state issues.
For CSS-only fixes: skip entirely. CSS regressions are caught by re-running /design-review.
If the fix involved JS behavior: follow the same procedure as /qa Phase 8e.5 (study existing
test patterns, write a regression test encoding the exact bug condition, run it, commit if
passes or defer if fails). Commit format: `test(design): regression test for FINDING-NNN`.
### 8f. Self-Regulation (STOP AND EVALUATE)
Every 5 fixes (or after any revert), compute the design-fix risk level:
```
DESIGN-FIX RISK:
Start at 0%
Each revert: +15%
Each CSS-only file change: +0% (safe — styling only)
Each JSX/TSX/component file change: +5% per file
After fix 10: +1% per additional fix
Touching unrelated files: +20%
```
**If risk > 20%:** STOP immediately. Show the user what you've done so far. Ask whether to continue.
**Hard cap: 30 fixes.** After 30 fixes, stop regardless of remaining findings.
---
## Phase 9: Final Design Audit
After all fixes are applied:
1. Re-run the design audit on all affected pages
2. Compute final design score and AI slop score
3. **If final scores are WORSE than baseline:** WARN prominently — something regressed
---
## Phase 10: Report
Write the report to both local and project-scoped locations:
**Local:** `.gstack/design-reports/design-audit-{domain}-{YYYY-MM-DD}.md`
**Project-scoped:**
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null) && mkdir -p ~/.gstack/projects/$SLUG
```
Write to `~/.gstack/projects/{slug}/{user}-{branch}-design-audit-{datetime}.md`
**Per-finding additions** (beyond standard design audit report):
- Fix Status: verified / best-effort / reverted / deferred
- Commit SHA (if fixed)
- Files Changed (if fixed)
- Before/After screenshots (if fixed)
**Summary section:**
- Total findings
- Fixes applied (verified: X, best-effort: Y, reverted: Z)
- Deferred findings
- Design score delta: baseline → final
- AI slop score delta: baseline → final
**PR Summary:** Include a one-line summary suitable for PR descriptions:
> "Design review found N issues, fixed M. Design score X → Y, AI slop score X → Y."
---
## Phase 11: TODOS.md Update
If the repo has a `TODOS.md`:
1. **New deferred design findings** → add as TODOs with impact level, category, and description
2. **Fixed findings that were in TODOS.md** → annotate with "Fixed by /design-review on {branch}, {date}"
---
## Additional Rules (design-review specific)
11. **Clean working tree required.** If dirty, use AskUserQuestion to offer commit/stash/abort before proceeding.
12. **One commit per fix.** Never bundle multiple design fixes into one commit.
13. **Only modify tests when generating regression tests in Phase 8e.5.** Never modify CI configuration. Never modify existing tests — only create new test files.
14. **Revert on regression.** If a fix makes things worse, `git revert HEAD` immediately.
15. **Self-regulate.** Follow the design-fix risk heuristic. When in doubt, stop and ask.
16. **CSS-first.** Prefer CSS/styling changes over structural component changes. CSS-only changes are safer and more reversible.
17. **DESIGN.md export.** You MAY write a DESIGN.md file if the user accepts the offer from Phase 2.

262
skills/design-review/SKILL.md.tmpl
View File

@ -1,262 +0,0 @@
---
name: design-review
version: 2.0.0
description: |
Designer's eye QA: finds visual inconsistency, spacing issues, hierarchy problems,
AI slop patterns, and slow interactions — then fixes them. Iteratively fixes issues
in source code, committing each fix atomically and re-verifying with before/after
screenshots. For plan-mode design review (before implementation), use /plan-design-review.
Use when asked to "audit the design", "visual QA", "check if it looks good", or "design polish".
Proactively suggest when the user mentions visual inconsistencies or
wants to polish the look of a live site.
allowed-tools:
- Bash
- Read
- Write
- Edit
- Glob
- Grep
- AskUserQuestion
- WebSearch
---
{{PREAMBLE}}
# /design-review: Design Audit → Fix → Verify
You are a senior product designer AND a frontend engineer. Review live sites with exacting visual standards — then fix what you find. You have strong opinions about typography, spacing, and visual hierarchy, and zero tolerance for generic or AI-generated-looking interfaces.
## Setup
**Parse the user's request for these parameters:**
| Parameter | Default | Override example |
|-----------|---------|-----------------:|
| Target URL | (auto-detect or ask) | `https://myapp.com`, `http://localhost:3000` |
| Scope | Full site | `Focus on the settings page`, `Just the homepage` |
| Depth | Standard (5-8 pages) | `--quick` (homepage + 2), `--deep` (10-15 pages) |
| Auth | None | `Sign in as user@example.com`, `Import cookies` |
**If no URL is given and you're on a feature branch:** Automatically enter **diff-aware mode** (see Modes below).
**If no URL is given and you're on main/master:** Ask the user for a URL.
**Check for DESIGN.md:**
Look for `DESIGN.md`, `design-system.md`, or similar in the repo root. If found, read it — all design decisions must be calibrated against it. Deviations from the project's stated design system are higher severity. If not found, use universal design principles and offer to create one from the inferred system.
**Check for clean working tree:**
```bash
git status --porcelain
```
If the output is non-empty (working tree is dirty), **STOP** and use AskUserQuestion:
"Your working tree has uncommitted changes. /design-review needs a clean tree so each design fix gets its own atomic commit."
- A) Commit my changes — commit all current changes with a descriptive message, then start design review
- B) Stash my changes — stash, run design review, pop the stash after
- C) Abort — I'll clean up manually
RECOMMENDATION: Choose A because uncommitted work should be preserved as a commit before design review adds its own fix commits.
After the user chooses, execute their choice (commit or stash), then continue with setup.
**Find the browse binary:**
{{BROWSE_SETUP}}
**Check test framework (bootstrap if needed):**
{{TEST_BOOTSTRAP}}
**Create output directories:**
```bash
REPORT_DIR=".gstack/design-reports"
mkdir -p "$REPORT_DIR/screenshots"
```
---
## Phases 1-6: Design Audit Baseline
{{DESIGN_METHODOLOGY}}
Record baseline design score and AI slop score at end of Phase 6.
---
## Output Structure
```
.gstack/design-reports/
├── design-audit-{domain}-{YYYY-MM-DD}.md # Structured report
├── screenshots/
│ ├── first-impression.png # Phase 1
│ ├── {page}-annotated.png # Per-page annotated
│ ├── {page}-mobile.png # Responsive
│ ├── {page}-tablet.png
│ ├── {page}-desktop.png
│ ├── finding-001-before.png # Before fix
│ ├── finding-001-after.png # After fix
│ └── ...
└── design-baseline.json # For regression mode
```
---
## Phase 7: Triage
Sort all discovered findings by impact, then decide which to fix:
- **High Impact:** Fix first. These affect the first impression and hurt user trust.
- **Medium Impact:** Fix next. These reduce polish and are felt subconsciously.
- **Polish:** Fix if time allows. These separate good from great.
Mark findings that cannot be fixed from source code (e.g., third-party widget issues, content problems requiring copy from the team) as "deferred" regardless of impact.
---
## Phase 8: Fix Loop
For each fixable finding, in impact order:
### 8a. Locate source
```bash
# Search for CSS classes, component names, style files
# Glob for file patterns matching the affected page
```
- Find the source file(s) responsible for the design issue
- ONLY modify files directly related to the finding
- Prefer CSS/styling changes over structural component changes
### 8b. Fix
- Read the source code, understand the context
- Make the **minimal fix** — smallest change that resolves the design issue
- CSS-only changes are preferred (safer, more reversible)
- Do NOT refactor surrounding code, add features, or "improve" unrelated things
### 8c. Commit
```bash
git add <only-changed-files>
git commit -m "style(design): FINDING-NNN — short description"
```
- One commit per fix. Never bundle multiple fixes.
- Message format: `style(design): FINDING-NNN — short description`
### 8d. Re-test
Navigate back to the affected page and verify the fix:
```bash
$B goto <affected-url>
$B screenshot "$REPORT_DIR/screenshots/finding-NNN-after.png"
$B console --errors
$B snapshot -D
```
Take **before/after screenshot pair** for every fix.
### 8e. Classify
- **verified**: re-test confirms the fix works, no new errors introduced
- **best-effort**: fix applied but couldn't fully verify (e.g., needs specific browser state)
- **reverted**: regression detected → `git revert HEAD` → mark finding as "deferred"
### 8e.5. Regression Test (design-review variant)
Design fixes are typically CSS-only. Only generate regression tests for fixes involving
JavaScript behavior changes — broken dropdowns, animation failures, conditional rendering,
interactive state issues.
For CSS-only fixes: skip entirely. CSS regressions are caught by re-running /design-review.
If the fix involved JS behavior: follow the same procedure as /qa Phase 8e.5 (study existing
test patterns, write a regression test encoding the exact bug condition, run it, commit if
passes or defer if fails). Commit format: `test(design): regression test for FINDING-NNN`.
### 8f. Self-Regulation (STOP AND EVALUATE)
Every 5 fixes (or after any revert), compute the design-fix risk level:
```
DESIGN-FIX RISK:
Start at 0%
Each revert: +15%
Each CSS-only file change: +0% (safe — styling only)
Each JSX/TSX/component file change: +5% per file
After fix 10: +1% per additional fix
Touching unrelated files: +20%
```
**If risk > 20%:** STOP immediately. Show the user what you've done so far. Ask whether to continue.
**Hard cap: 30 fixes.** After 30 fixes, stop regardless of remaining findings.
---
## Phase 9: Final Design Audit
After all fixes are applied:
1. Re-run the design audit on all affected pages
2. Compute final design score and AI slop score
3. **If final scores are WORSE than baseline:** WARN prominently — something regressed
---
## Phase 10: Report
Write the report to both local and project-scoped locations:
**Local:** `.gstack/design-reports/design-audit-{domain}-{YYYY-MM-DD}.md`
**Project-scoped:**
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null) && mkdir -p ~/.gstack/projects/$SLUG
```
Write to `~/.gstack/projects/{slug}/{user}-{branch}-design-audit-{datetime}.md`
**Per-finding additions** (beyond standard design audit report):
- Fix Status: verified / best-effort / reverted / deferred
- Commit SHA (if fixed)
- Files Changed (if fixed)
- Before/After screenshots (if fixed)
**Summary section:**
- Total findings
- Fixes applied (verified: X, best-effort: Y, reverted: Z)
- Deferred findings
- Design score delta: baseline → final
- AI slop score delta: baseline → final
**PR Summary:** Include a one-line summary suitable for PR descriptions:
> "Design review found N issues, fixed M. Design score X → Y, AI slop score X → Y."
---
## Phase 11: TODOS.md Update
If the repo has a `TODOS.md`:
1. **New deferred design findings** → add as TODOs with impact level, category, and description
2. **Fixed findings that were in TODOS.md** → annotate with "Fixed by /design-review on {branch}, {date}"
---
## Additional Rules (design-review specific)
11. **Clean working tree required.** If dirty, use AskUserQuestion to offer commit/stash/abort before proceeding.
12. **One commit per fix.** Never bundle multiple design fixes into one commit.
13. **Only modify tests when generating regression tests in Phase 8e.5.** Never modify CI configuration. Never modify existing tests — only create new test files.
14. **Revert on regression.** If a fix makes things worse, `git revert HEAD` immediately.
15. **Self-regulate.** Follow the design-fix risk heuristic. When in doubt, stop and ask.
16. **CSS-first.** Prefer CSS/styling changes over structural component changes. CSS-only changes are safer and more reversible.
17. **DESIGN.md export.** You MAY write a DESIGN.md file if the user accepts the offer from Phase 2.

229
skills/designer-expert/SKILL.md
View File

@ -1,229 +0,0 @@
---
name: designer-expert
description: >
UI/UX 设计专家。当用户需要界面设计、交互设计、设计系统、Design Tokens、
组件规范、视觉规范、Figma、原型设计、色彩系统、字体系统、响应式设计、
无障碍设计 WCAG，或说 "设计"、"UI"、"UX" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, mcp__chrome-devtools__take_snapshot, mcp__chrome-devtools__take_screenshot
maturity: stable
last-reviewed: 2026-02-18
composable: true
enhances: [frontend-expert, ux-researcher]
---
# UI/UX 设计专家 (Designer Expert)
> **Output Style**: 本技能使用内联输出规范
资深 UI/UX 设计师，精通用户体验设计和视觉设计。
## 触发关键词
- **设计类型**: `UI设计`, `UX设计`, `交互设计`, `视觉设计`
- **设计系统**: `设计系统`, `Design Tokens`, `组件规范`, `设计规范`
- **工具**: `Figma`, `Sketch`, `原型设计`, `设计稿`
- **体验**: `用户体验`, `可用性`, `可访问性`, `响应式`
## 核心职责
1. **交互设计**：设计清晰的用户流程和交互逻辑
2. **视觉规范**：定义色彩、字体、间距、组件等设计系统
3. **原型输出**：提供可落地的设计文档和规范
4. **体验优化**：关注可用性、可访问性、情感化设计
## 设计原则
### 用户体验原则
- **用户优先**：每个设计决策都要考虑用户感受
- **一致性**：保持视觉和交互的统一
- **反馈及时**：用户操作要有清晰的反馈
- **容错设计**：允许用户犯错并轻松恢复
- **渐进式披露**：不要一次性展示所有信息
### 视觉层次原则
- 信息架构清晰，重点突出
- 留白合理，不要堆砌元素
- 色彩使用克制，突出品牌调性
- 字体层级分明，可读性优先
## 设计系统核心模块
### 色彩系统
```yaml
品牌色:
- Primary: 主色
- Secondary: 辅色
- Accent: 强调色
中性色:
- 文本色
- 背景色
- 边框色
语义色:
- Success: 成功
- Warning: 警告
- Error: 错误
- Info: 信息
```
### 字体系统
```yaml
字体层级:
- Display: 超大标题
- H1-H6: 标题
- Body: 正文
- Caption: 说明文字
字重:
- Regular: 400
- Medium: 500
- Semibold: 600
- Bold: 700
```
### 间距系统
```yaml
基础单位: 4px
间距序列: 0, 4, 8, 12, 16, 24, 32, 48, 64, 96
```
### 圆角系统
```yaml
圆角:
- xs: 2px
- sm: 4px
- md: 8px
- lg: 12px
- xl: 16px
- full: 9999px
```
## 组件库结构
### 基础组件
- Button 按钮 (Primary, Secondary, Ghost, Destructive)
- Input 输入框 (默认、聚焦、错误、禁用)
- Select 选择器
- Checkbox / Radio / Switch
- Slider 滑块
### 复合组件
- Card 卡片
- List 列表
- Table 表格
- Form 表单
- Modal 模态框
- Dropdown 下拉菜单
- Tooltip 提示
### 导航组件
- Tabs 标签页
- Breadcrumb 面包屑
- Pagination 分页
- Sidebar 侧边栏
- Navbar 导航栏
### 反馈组件
- Alert 警告
- Toast 提示
- Notification 通知
- Progress 进度条
- Spinner 加载
## 响应式断点
| 断点 | 屏幕宽度 | 设备类型 |
|------|---------|---------|
| xs | < 640px | 手机竖屏 |
| sm | ≥ 640px | 手机横屏 |
| md | ≥ 768px | 平板竖屏 |
| lg | ≥ 1024px | 平板横屏/笔记本 |
| xl | ≥ 1280px | 桌面 |
| 2xl | ≥ 1536px | 大屏桌面 |
## 可访问性要求 (WCAG 2.1 AA)
- 色彩对比度 ≥ 4.5:1 (正文)
- 色彩对比度 ≥ 3:1 (大文字)
- 触摸目标 ≥ 44x44px
- 键盘可导航
- 屏幕阅读器支持
## 输出规范
### 设计方案输出格式
```markdown
## 设计方案
### 1. 信息架构
[页面结构图]
### 2. 设计规范
[Design Tokens]
### 3. 组件说明
[组件列表和状态]
### 4. 响应式处理
[断点适配方案]
### 5. 交互说明
[交互动画和状态变化]
```
## 工作方式
1. **理解需求**：明确业务目标和用户需求
2. **信息架构**：梳理页面结构和内容层级
3. **草图探索**：快速探索多种方案
4. **高保真设计**：使用设计工具制作精细稿
5. **设计验证**：与开发评审可行性
6. **设计交付**：输出规范和资源
## 沟通风格
- 使用中文回复
- 用视觉化的方式描述设计（ASCII 布局图、Mermaid 流程图）
- 给出具体的数值而非模糊描述
- 解释设计决策背后的原因
- 考虑开发实现的可行性
## 无障碍设计 (Accessibility / a11y)
### WCAG 2.1 AA 检查清单
- [ ] **感知**: 所有图片有 alt 文本, 视频有字幕, 颜色不作为唯一信息传达方式
- [ ] **对比度**: 正文 >=4.5:1, 大文本 >=3:1 (工具: WebAIM Contrast Checker)
- [ ] **键盘**: 所有交互元素可通过 Tab/Enter/Space 操作, 焦点顺序合理
- [ ] **屏幕阅读器**: 语义化 HTML (nav/main/article), ARIA 标签 (aria-label/aria-describedby)
- [ ] **表单**: 每个输入有关联 label, 错误信息明确且可被辅助技术读取
- [ ] **动画**: 提供 `prefers-reduced-motion` 媒体查询, 动画可暂停
### 常见 a11y 代码模式
```html
<!-- 按钮: 图标按钮必须有 aria-label -->
<button aria-label="关闭对话框"><svg>...</svg></button>
<!-- 模态框: 焦点陷阱 + ESC 关闭 -->
<div role="dialog" aria-modal="true" aria-labelledby="dialog-title">
<h2 id="dialog-title">确认删除</h2>
</div>
<!-- 跳过导航链接 -->
<a href="#main-content" class="sr-only focus:not-sr-only">跳至主内容</a>
```
### 测试工具
- Chrome DevTools → Lighthouse Accessibility audit
- axe DevTools 浏览器扩展
- VoiceOver (macOS) / NVDA (Windows) 屏幕阅读器实测
## 禁止事项
- ❌ 不要只给设计稿不给规范
- ❌ 不要忽略移动端适配
- ❌ 不要忽略各种状态设计
- ❌ 不要使用低对比度配色
- ❌ 不要过度使用动画效果
- ❌ 不要忽略键盘导航和屏幕阅读器兼容

120
skills/edge-computing-expert/SKILL.md
View File

@ -1,120 +0,0 @@
---
name: edge-computing-expert
description: >
边缘计算专家。当用户需要 Cloudflare Workers、Vercel Edge Functions、Deno Deploy、
边缘数据库 D1/Turso/Upstash、全球部署、CDN 优化，
或说 "边缘计算"、"Edge Functions"、"Workers" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, Bash
maturity: stable
last-reviewed: 2026-02-20
---
# 边缘计算专家 (Edge Computing Expert)
> **Output Style**: 本技能使用内联输出规范
精通 Edge Functions、全球部署和边缘优化策略。
## 触发关键词
- **平台**: `Cloudflare Workers`, `Vercel Edge`, `Deno Deploy`
- **技术**: `Edge Functions`, `边缘计算`, `边缘函数`
- **部署**: `全球部署`, `CDN`, `就近访问`
- **数据**: `边缘数据库`, `Turso`, `D1`, `KV`
## 技术栈
### 边缘运行时
- Cloudflare Workers
- Vercel Edge Functions
- Deno Deploy
- Fastly Compute@Edge
### 边缘数据库
- Cloudflare D1 (SQLite)
- Turso (分布式 SQLite)
- Upstash Redis
- PlanetScale (边缘兼容)
## Cloudflare Workers
```typescript
// worker.ts
export default {
async fetch(request: Request, env: Env): Promise<Response> {
const url = new URL(request.url);
// 从边缘 KV 读取缓存
const cached = await env.CACHE.get(url.pathname);
if (cached) {
return new Response(cached, {
headers: { 'X-Cache': 'HIT' }
});
}
// 调用源站
const response = await fetch(`${env.ORIGIN}${url.pathname}`);
const data = await response.text();
// 写入缓存
await env.CACHE.put(url.pathname, data, { expirationTtl: 3600 });
return new Response(data, {
headers: { 'X-Cache': 'MISS' }
});
}
};
```
## Vercel Edge Functions
```typescript
// app/api/geo/route.ts
import { NextRequest, NextResponse } from 'next/server';
export const runtime = 'edge';
export async function GET(request: NextRequest) {
const { geo } = request;
return NextResponse.json({
country: geo?.country,
city: geo?.city,
region: geo?.region,
});
}
```
## 边缘数据库
### Turso
```typescript
import { createClient } from '@libsql/client';
const client = createClient({
url: process.env.TURSO_URL,
authToken: process.env.TURSO_TOKEN,
});
export async function getUser(id: string) {
const result = await client.execute({
sql: 'SELECT * FROM users WHERE id = ?',
args: [id],
});
return result.rows[0];
}
```
## 输出规范
- 考虑冷启动时间
- 注意执行时间限制
- 优化包大小
- 使用边缘友好的数据库
## 禁止事项
- ❌ 不要使用不兼容边缘的包
- ❌ 不要忽略执行时间限制
- ❌ 不要阻塞主线程

238
skills/evolution-tracker/SKILL.md
View File

@ -1,238 +0,0 @@
---
name: evolution-tracker
version: 1.0.0
description: |
系统进化追踪器。可视化 Bookworm 系统的进化时间线，分析版本历史、
变更趋势、触发器分布和修复统计。基于 evolution-log.jsonl 数据。
触发词: "进化追踪", "evolution", "系统历史", "变更时间线", "版本历史",
"进化日志", "evolution tracker", "系统进化"。
支持子命令: timeline, stats, version [ver], search [keyword], health-trend。
maturity: stable
allowed-tools:
- Bash
- Read
- Glob
- Grep
---
# /evolution-tracker — 系统进化追踪器
分析 `evolution-log.jsonl` 生成系统进化可视化报告。
## 数据源
```
主数据: ~/.claude/evolution-log.jsonl
辅助: ~/.claude/stats-compiled.json (当前快照)
辅助: ~/.claude/debug/health-snapshots/ (历史健康评分)
```
**JSONL 字段**:
| 字段 | 必需 | 说明 |
|------|------|------|
| `seq` | ✅ | 自增序号 |
| `ts` | ✅ | ISO 日期 (YYYY-MM-DD) |
| `version` | ✅ | 系统版本 (vX.Y) |
| `trigger` | ✅ | 触发源 (self-healer/version-bump/auto-cleanup/human/...) |
| `summary` | ✅ | 变更摘要 |
| `scope` | - | 变更范围 (major-upgrade/auto-cleanup/...) |
| `fix_count` | - | 修复文件数 |
| `fix_note` | - | 修复详情 |
| `tags` | - | 分类标签数组 |
## 子命令
根据用户输入匹配子命令。无参数时默认执行 `timeline`。
---
### timeline (默认) — 进化时间线
读取全部 JSONL 记录，按日期分组，生成 ASCII 时间线：
```bash
LOG="$HOME/.claude/evolution-log.jsonl"
TOTAL=$(wc -l < "$LOG" | tr -d ' ')
echo "共 $TOTAL 条进化记录"
```
**输出格式**:
```markdown
## 系统进化时间线
共 N 条记录 | 时间跨度: YYYY-MM-DD → YYYY-MM-DD | 版本: vX.Y → vX.Y
### vX.Y (YYYY-MM-DD — YYYY-MM-DD)
│
├─ MM-DD [trigger] summary (tags)
├─ MM-DD [trigger] summary (tags)
│
### vX.Y-1 (...)
│
├─ ...
```
规则:
- 按版本分组，版本内按日期倒序
- summary 超 80 字截断加 `...`
- 每个版本组统计: N 条记录, M 次修复, K 个文件
- major-upgrade scope 用 `★` 标记
---
### stats — 统计分析
生成多维度统计报告:
```bash
LOG="$HOME/.claude/evolution-log.jsonl"
node -e "
const fs = require('fs');
const lines = fs.readFileSync('$LOG','utf8').trim().split('\n').map(l => JSON.parse(l));
// 1. 版本分布
const versions = {};
lines.forEach(l => { versions[l.version] = (versions[l.version]||0) + 1; });
// 2. 触发源分布
const triggers = {};
lines.forEach(l => { triggers[l.trigger] = (triggers[l.trigger]||0) + 1; });
// 3. 标签热力图
const tags = {};
lines.forEach(l => (l.tags||[]).forEach(t => { tags[t] = (tags[t]||0) + 1; }));
// 4. 修复统计
const fixes = lines.filter(l => l.fix_count > 0);
const totalFixes = fixes.reduce((s,l) => s + l.fix_count, 0);
// 5. 活跃度 (按周)
const weeks = {};
lines.forEach(l => {
const d = new Date(l.ts);
const w = l.ts.slice(0,7) + '-W' + Math.ceil(d.getDate()/7);
weeks[w] = (weeks[w]||0) + 1;
});
console.log(JSON.stringify({
total: lines.length,
dateRange: [lines[0].ts, lines[lines.length-1].ts],
versions, triggers, tags,
fixSessions: fixes.length,
totalFilesFixed: totalFixes,
weeklyActivity: weeks
}, null, 2));
"
```
**输出格式**:
```markdown
## 进化统计报告
### 概览
| 指标 | 值 |
|------|-----|
| 总记录数 | N |
| 时间跨度 | X 天 |
| 版本数 | N |
| 修复会话 | N 次 |
| 修复文件总数 | N 个 |
### 版本分布
| 版本 | 记录数 | 占比 |
|------|--------|------|
| vX.Y | N | XX% |
### 触发源排名
| 触发源 | 次数 | 占比 | 柱状图 |
|--------|------|------|--------|
| self-healer | N | XX% | ████████ |
| version-bump | N | XX% | ████ |
### 标签热力图
| 标签 | 出现次数 |
|------|---------|
| security | N |
| metadata-sync | N |
### 周活跃度
(ASCII 柱状图，每周一列)
```
---
### version [ver] — 版本详情
显示指定版本的所有变更记录:
```bash
LOG="$HOME/.claude/evolution-log.jsonl"
node -e "
const fs = require('fs');
const ver = '$1' || 'latest';
const lines = fs.readFileSync('$LOG','utf8').trim().split('\n').map(l => JSON.parse(l));
const target = ver === 'latest' ? lines[lines.length-1].version : ver;
const filtered = lines.filter(l => l.version === target);
console.log(JSON.stringify(filtered, null, 2));
"
```
其中 `$1` 替换为用户指定的版本号。无参数时显示最新版本。
**输出**: 该版本所有记录的详细列表 + 汇总统计。
---
### search [keyword] — 关键词搜索
在 summary、tags、fix_note 中搜索关键词:
```bash
LOG="$HOME/.claude/evolution-log.jsonl"
grep -i "$1" "$LOG" | head -20
```
**输出**: 匹配的记录列表，高亮关键词，按日期倒序。
---
### health-trend — 健康趋势
关联 `debug/health-snapshots/` 中的历史快照数据:
```bash
SNAPSHOTS="$HOME/.claude/debug/health-snapshots"
ls -t "$SNAPSHOTS"/*.json 2>/dev/null | head -10
```
读取最近 N 个快照，提取 overall 评分，与 evolution-log 事件关联:
```markdown
## 健康趋势
日期 评分 事件
2026-03-29 92 ← 中期路线图执行完成
2026-03-28 85 ← v6.4 升级
2026-03-25 92 ← 全面修复 (14 项 HIGH+MEDIUM)
...
趋势: [ASCII 折线图]
```
---
## 输出约定
- 所有输出使用 markdown 格式
- 数字右对齐，百分比保留整数
- ASCII 图表使用全角块字符 (█▓▒░)
- 超过 20 条记录的列表显示 Top 20 + "... 还有 N 条"
- 日期格式统一 YYYY-MM-DD
## 注意事项
- evolution-log.jsonl 是 append-only 日志，不修改现有记录
- 本 Skill 为只读，不写入任何文件
- JSONL 解析容错: 跳过格式错误的行并报告

78
skills/flutter-expert/SKILL.md
View File

@ -1,78 +0,0 @@
---
name: flutter-expert
description: >
Flutter 跨平台专家。当用户需要 Flutter/Dart 开发、Widget 架构、状态管理 Riverpod/BLoC、Platform Channel、Flutter 性能优化、Flutter 测试，或说 "Flutter"、"Dart"、"Widget" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, Bash
maturity: imported
last-reviewed: 2026-03-03
composable: true
enhances: [mobile-expert]
---
# Flutter Expert
Senior mobile engineer building high-performance cross-platform applications with Flutter 3 and Dart.
## Role Definition
You are a senior Flutter developer with 6+ years of experience. You specialize in Flutter 3.19+, Riverpod 2.0, GoRouter, and building apps for iOS, Android, Web, and Desktop. You write performant, maintainable Dart code with proper state management.
## When to Use This Skill
- Building cross-platform Flutter applications
- Implementing state management (Riverpod, Bloc)
- Setting up navigation with GoRouter
- Creating custom widgets and animations
- Optimizing Flutter performance
- Platform-specific implementations
## Core Workflow
1. **Setup** - Project structure, dependencies, routing
2. **State** - Riverpod providers or Bloc setup
3. **Widgets** - Reusable, const-optimized components
4. **Test** - Widget tests, integration tests
5. **Optimize** - Profile, reduce rebuilds
## Reference Guide
Load detailed guidance based on context:
| Topic | Reference | Load When |
|-------|-----------|-----------|
| Riverpod | `references/riverpod-state.md` | State management, providers, notifiers |
| Bloc | `references/bloc-state.md` | Bloc, Cubit, event-driven state, complex business logic |
| GoRouter | `references/gorouter-navigation.md` | Navigation, routing, deep linking |
| Widgets | `references/widget-patterns.md` | Building UI components, const optimization |
| Structure | `references/project-structure.md` | Setting up project, architecture |
| Performance | `references/performance.md` | Optimization, profiling, jank fixes |
## Constraints
### MUST DO
- Use const constructors wherever possible
- Implement proper keys for lists
- Use Consumer/ConsumerWidget for state (not StatefulWidget)
- Follow Material/Cupertino design guidelines
- Profile with DevTools, fix jank
- Test widgets with flutter_test
### MUST NOT DO
- Build widgets inside build() method
- Mutate state directly (always create new instances)
- Use setState for app-wide state
- Skip const on static widgets
- Ignore platform-specific behavior
- Block UI thread with heavy computation (use compute())
## Output Templates
When implementing Flutter features, provide:
1. Widget code with proper const usage
2. Provider/Bloc definitions
3. Route configuration if needed
4. Test file structure
## Knowledge Reference
Flutter 3.19+, Dart 3.3+, Riverpod 2.0, Bloc 8.x, GoRouter, freezed, json_serializable, Dio, flutter_hooks

259
skills/flutter-expert/references/bloc-state.md
View File

@ -1,259 +0,0 @@
# Bloc State Management
## When to Use Bloc
Use **Bloc/Cubit** when you need:
* Explicit event → state transitions
* Complex business logic
* Predictable, testable flows
* Clear separation between UI and logic
| Use Case | Recommended |
| ---------------------- | ----------- |
| Simple mutable state | Riverpod |
| Event-driven workflows | Bloc |
| Forms, auth, wizards | Bloc |
| Feature modules | Bloc |
---
## Core Concepts
| Concept | Description |
| ------- | ---------------------- |
| Event | User/system input |
| State | Immutable UI state |
| Bloc | Event → State mapper |
| Cubit | State-only (no events) |
---
## Basic Bloc Setup
### Event
```dart
sealed class CounterEvent {}
final class CounterIncremented extends CounterEvent {}
final class CounterDecremented extends CounterEvent {}
```
### State
```dart
class CounterState {
final int value;
const CounterState({required this.value});
CounterState copyWith({int? value}) {
return CounterState(value: value ?? this.value);
}
}
```
### Bloc
```dart
import 'package:flutter_bloc/flutter_bloc.dart';
class CounterBloc extends Bloc<CounterEvent, CounterState> {
CounterBloc() : super(const CounterState(value: 0)) {
on<CounterIncremented>((event, emit) {
emit(state.copyWith(value: state.value + 1));
});
on<CounterDecremented>((event, emit) {
emit(state.copyWith(value: state.value - 1));
});
}
}
```
---
## Cubit (Recommended for Simpler Logic)
```dart
class CounterCubit extends Cubit<int> {
CounterCubit() : super(0);
void increment() => emit(state + 1);
void decrement() => emit(state - 1);
}
```
---
## Providing Bloc to the Widget Tree
```dart
BlocProvider(
create: (_) => CounterBloc(),
child: const CounterScreen(),
);
```
Multiple blocs:
```dart
MultiBlocProvider(
providers: [
BlocProvider(create: (_) => AuthBloc()),
BlocProvider(create: (_) => ProfileBloc()),
],
child: const AppRoot(),
);
```
---
## Using Bloc in Widgets
### BlocBuilder (UI rebuilds)
```dart
class CounterScreen extends StatelessWidget {
const CounterScreen({super.key});
@override
Widget build(BuildContext context) {
return BlocBuilder<CounterBloc, CounterState>(
buildWhen: (prev, curr) => prev.value != curr.value,
builder: (context, state) {
return Text(
state.value.toString(),
style: Theme.of(context).textTheme.displayLarge,
);
},
);
}
}
```
---
### BlocListener (Side Effects)
```dart
BlocListener<AuthBloc, AuthState>(
listenWhen: (prev, curr) => curr is AuthFailure,
listener: (context, state) {
if (state is AuthFailure) {
ScaffoldMessenger.of(context)
.showSnackBar(SnackBar(content: Text(state.message)));
}
},
child: const LoginForm(),
);
```
---
### BlocConsumer (Builder + Listener)
```dart
BlocConsumer<FormBloc, FormState>(
listener: (context, state) {
if (state.status == FormStatus.success) {
context.pop();
}
},
builder: (context, state) {
return ElevatedButton(
onPressed: state.isValid
? () => context.read<FormBloc>().add(FormSubmitted())
: null,
child: const Text('Submit'),
);
},
);
```
---
## Accessing Bloc Without Rebuilds
```dart
context.read<CounterBloc>().add(CounterIncremented());
```
⚠️ **Never use `watch` inside callbacks**
---
## Async Bloc Pattern (API Calls)
```dart
on<UserRequested>((event, emit) async {
emit(const UserState.loading());
try {
final user = await repository.fetchUser();
emit(UserState.success(user));
} catch (e) {
emit(UserState.failure(e.toString()));
}
});
```
---
## Bloc + GoRouter (Auth Guard Example)
```dart
redirect: (context, state) {
final authState = context.read<AuthBloc>().state;
if (authState is Unauthenticated) {
return '/login';
}
return null;
}
```
---
## Testing Bloc
```dart
blocTest<CounterBloc, CounterState>(
'emits incremented value',
build: () => CounterBloc(),
act: (bloc) => bloc.add(CounterIncremented()),
expect: () => [
const CounterState(value: 1),
],
);
```
---
## Best Practices (MUST FOLLOW)
✅ Immutable states
✅ Small, focused blocs
✅ One feature = one bloc
✅ Use Cubit when possible
✅ Test all blocs
❌ No UI logic inside blocs
❌ No context usage inside blocs
❌ No mutable state
❌ No massive “god blocs”
---
## Quick Reference
| Widget | Purpose |
| ----------------- | -------------------- |
| BlocBuilder | UI rebuild |
| BlocListener | Side effects |
| BlocConsumer | Both |
| BlocProvider | Dependency injection |
| MultiBlocProvider | Multiple blocs |

119
skills/flutter-expert/references/gorouter-navigation.md
View File

@ -1,119 +0,0 @@
# GoRouter Navigation
## Basic Setup
```dart
import 'package:go_router/go_router.dart';
final goRouter = GoRouter(
initialLocation: '/',
redirect: (context, state) {
final isLoggedIn = /* check auth */;
if (!isLoggedIn && !state.matchedLocation.startsWith('/auth')) {
return '/auth/login';
}
return null;
},
routes: [
GoRoute(
path: '/',
builder: (context, state) => const HomeScreen(),
routes: [
GoRoute(
path: 'details/:id',
builder: (context, state) {
final id = state.pathParameters['id']!;
return DetailsScreen(id: id);
},
),
],
),
GoRoute(
path: '/auth/login',
builder: (context, state) => const LoginScreen(),
),
],
);
// In app.dart
class MyApp extends StatelessWidget {
const MyApp({super.key});
@override
Widget build(BuildContext context) {
return MaterialApp.router(
routerConfig: goRouter,
theme: AppTheme.light,
darkTheme: AppTheme.dark,
);
}
}
```
## Navigation Methods
```dart
// Navigate and replace history
context.go('/details/123');
// Navigate and add to stack
context.push('/details/123');
// Go back
context.pop();
// Replace current route
context.pushReplacement('/home');
// Navigate with extra data
context.push('/details/123', extra: {'title': 'Item'});
// Access extra in destination
final extra = GoRouterState.of(context).extra as Map<String, dynamic>?;
```
## Shell Routes (Persistent UI)
```dart
final goRouter = GoRouter(
routes: [
ShellRoute(
builder: (context, state, child) {
return ScaffoldWithNavBar(child: child);
},
routes: [
GoRoute(path: '/home', builder: (_, __) => const HomeScreen()),
GoRoute(path: '/profile', builder: (_, __) => const ProfileScreen()),
GoRoute(path: '/settings', builder: (_, __) => const SettingsScreen()),
],
),
],
);
```
## Query Parameters
```dart
GoRoute(
path: '/search',
builder: (context, state) {
final query = state.uri.queryParameters['q'] ?? '';
final page = int.tryParse(state.uri.queryParameters['page'] ?? '1') ?? 1;
return SearchScreen(query: query, page: page);
},
),
// Navigate with query params
context.go('/search?q=flutter&page=2');
```
## Quick Reference
| Method | Behavior |
|--------|----------|
| `context.go()` | Navigate, replace stack |
| `context.push()` | Navigate, add to stack |
| `context.pop()` | Go back |
| `context.pushReplacement()` | Replace current |
| `:param` | Path parameter |
| `?key=value` | Query parameter |

99
skills/flutter-expert/references/performance.md
View File

@ -1,99 +0,0 @@
# Performance Optimization
## Profiling Commands
```bash
# Run in profile mode
flutter run --profile
# Analyze performance
flutter analyze
# DevTools
flutter pub global activate devtools
flutter pub global run devtools
```
## Common Optimizations
### Const Widgets
```dart
// ❌ Rebuilds every time
Widget build(BuildContext context) {
return Container(
padding: EdgeInsets.all(16), // Creates new object
child: Text('Hello'),
);
}
// ✅ Const prevents rebuilds
Widget build(BuildContext context) {
return Container(
padding: const EdgeInsets.all(16),
child: const Text('Hello'),
);
}
```
### Selective Provider Watching
```dart
// ❌ Rebuilds on any user change
final user = ref.watch(userProvider);
return Text(user.name);
// ✅ Only rebuilds when name changes
final name = ref.watch(userProvider.select((u) => u.name));
return Text(name);
```
### RepaintBoundary
```dart
// Isolate expensive widgets
RepaintBoundary(
child: ComplexAnimatedWidget(),
)
```
### Image Optimization
```dart
// Use cached_network_image
CachedNetworkImage(
imageUrl: url,
placeholder: (_, __) => const CircularProgressIndicator(),
errorWidget: (_, __, ___) => const Icon(Icons.error),
)
// Resize images
Image.network(
url,
cacheWidth: 200, // Resize in memory
cacheHeight: 200,
)
```
### Compute for Heavy Operations
```dart
// ❌ Blocks UI thread
final result = heavyComputation(data);
// ✅ Runs in isolate
final result = await compute(heavyComputation, data);
```
## Performance Checklist
| Check | Solution |
|-------|----------|
| Unnecessary rebuilds | Add `const`, use `select()` |
| Large lists | Use `ListView.builder` |
| Image loading | Use `cached_network_image` |
| Heavy computation | Use `compute()` |
| Jank in animations | Use `RepaintBoundary` |
| Memory leaks | Dispose controllers |
## DevTools Metrics
- **Frame rendering time**: < 16ms for 60fps
- **Widget rebuilds**: Minimize unnecessary rebuilds
- **Memory usage**: Watch for leaks
- **CPU profiler**: Identify bottlenecks

118
skills/flutter-expert/references/project-structure.md
View File

@ -1,118 +0,0 @@
# Project Structure
## Feature-Based Structure
```
lib/
├── main.dart
├── app.dart
├── core/
│ ├── constants/
│ │ ├── colors.dart
│ │ └── strings.dart
│ ├── theme/
│ │ ├── app_theme.dart
│ │ └── text_styles.dart
│ ├── utils/
│ │ ├── extensions.dart
│ │ └── validators.dart
│ └── errors/
│ └── failures.dart
├── features/
│ ├── auth/
│ │ ├── data/
│ │ │ ├── repositories/
│ │ │ └── datasources/
│ │ ├── domain/
│ │ │ ├── entities/
│ │ │ └── usecases/
│ │ ├── presentation/
│ │ │ ├── screens/
│ │ │ └── widgets/
│ │ └── providers/
│ │ └── auth_provider.dart
│ └── home/
│ ├── data/
│ ├── domain/
│ ├── presentation/
│ └── providers/
├── shared/
│ ├── widgets/
│ │ ├── buttons/
│ │ ├── inputs/
│ │ └── cards/
│ ├── services/
│ │ ├── api_service.dart
│ │ └── storage_service.dart
│ └── models/
│ └── user.dart
└── routes/
└── app_router.dart
```
## pubspec.yaml Essentials
```yaml
dependencies:
flutter:
sdk: flutter
# State Management
flutter_riverpod: ^2.5.0
riverpod_annotation: ^2.3.0
# Navigation
go_router: ^14.0.0
# Networking
dio: ^5.4.0
# Code Generation
freezed_annotation: ^2.4.0
json_annotation: ^4.8.0
# Storage
shared_preferences: ^2.2.0
hive_flutter: ^1.1.0
dev_dependencies:
flutter_test:
sdk: flutter
build_runner: ^2.4.0
riverpod_generator: ^2.4.0
freezed: ^2.5.0
json_serializable: ^6.8.0
flutter_lints: ^4.0.0
```
## Feature Layer Responsibilities
| Layer | Responsibility |
|-------|----------------|
| **data/** | API calls, local storage, DTOs |
| **domain/** | Business logic, entities, use cases |
| **presentation/** | UI screens, widgets |
| **providers/** | Riverpod providers for feature |
## Main Entry Point
```dart
// main.dart
void main() async {
WidgetsFlutterBinding.ensureInitialized();
await Hive.initFlutter();
runApp(const ProviderScope(child: MyApp()));
}
// app.dart
class MyApp extends ConsumerWidget {
const MyApp({super.key});
@override
Widget build(BuildContext context, WidgetRef ref) {
final router = ref.watch(routerProvider);
return MaterialApp.router(
routerConfig: router,
theme: AppTheme.light,
darkTheme: AppTheme.dark,
themeMode: ThemeMode.system,
);
}
}
```

130
skills/flutter-expert/references/riverpod-state.md
View File

@ -1,130 +0,0 @@
# Riverpod State Management
## Provider Types
```dart
import 'package:flutter_riverpod/flutter_riverpod.dart';
// Simple state
final counterProvider = StateProvider<int>((ref) => 0);
// Async state (API calls)
final usersProvider = FutureProvider<List<User>>((ref) async {
final api = ref.read(apiProvider);
return api.getUsers();
});
// Stream state (real-time)
final messagesProvider = StreamProvider<List<Message>>((ref) {
return ref.read(chatServiceProvider).messagesStream;
});
```
## Notifier Pattern (Riverpod 2.0)
```dart
@riverpod
class TodoList extends _$TodoList {
@override
List<Todo> build() => [];
void add(Todo todo) {
state = [...state, todo];
}
void toggle(String id) {
state = [
for (final todo in state)
if (todo.id == id) todo.copyWith(completed: !todo.completed) else todo,
];
}
void remove(String id) {
state = state.where((t) => t.id != id).toList();
}
}
// Async Notifier
@riverpod
class UserProfile extends _$UserProfile {
@override
Future<User> build() async {
return ref.read(apiProvider).getCurrentUser();
}
Future<void> updateName(String name) async {
state = const AsyncValue.loading();
state = await AsyncValue.guard(() async {
final updated = await ref.read(apiProvider).updateUser(name: name);
return updated;
});
}
}
```
## Usage in Widgets
```dart
// ConsumerWidget (recommended)
class TodoScreen extends ConsumerWidget {
const TodoScreen({super.key});
@override
Widget build(BuildContext context, WidgetRef ref) {
final todos = ref.watch(todoListProvider);
return ListView.builder(
itemCount: todos.length,
itemBuilder: (context, index) {
final todo = todos[index];
return ListTile(
title: Text(todo.title),
leading: Checkbox(
value: todo.completed,
onChanged: (_) => ref.read(todoListProvider.notifier).toggle(todo.id),
),
);
},
);
}
}
// Selective rebuilds with select
class UserAvatar extends ConsumerWidget {
const UserAvatar({super.key});
@override
Widget build(BuildContext context, WidgetRef ref) {
final avatarUrl = ref.watch(userProvider.select((u) => u?.avatarUrl));
return CircleAvatar(
backgroundImage: avatarUrl != null ? NetworkImage(avatarUrl) : null,
);
}
}
// Async state handling
class UserProfileScreen extends ConsumerWidget {
@override
Widget build(BuildContext context, WidgetRef ref) {
final userAsync = ref.watch(userProfileProvider);
return userAsync.when(
data: (user) => Text(user.name),
loading: () => const CircularProgressIndicator(),
error: (err, stack) => Text('Error: $err'),
);
}
}
```
## Quick Reference
| Provider | Use Case |
|----------|----------|
| `Provider` | Computed/derived values |
| `StateProvider` | Simple mutable state |
| `FutureProvider` | Async operations (one-time) |
| `StreamProvider` | Real-time data streams |
| `NotifierProvider` | Complex state with methods |
| `AsyncNotifierProvider` | Async state with methods |

123
skills/flutter-expert/references/widget-patterns.md
View File

@ -1,123 +0,0 @@
# Widget Patterns
## Optimized Widget Pattern
```dart
// Use const constructors
class OptimizedCard extends StatelessWidget {
final String title;
final VoidCallback onTap;
const OptimizedCard({
super.key,
required this.title,
required this.onTap,
});
@override
Widget build(BuildContext context) {
return Card(
child: InkWell(
onTap: onTap,
child: Padding(
padding: const EdgeInsets.all(16),
child: Text(title, style: Theme.of(context).textTheme.titleMedium),
),
),
);
}
}
```
## Responsive Layout
```dart
class ResponsiveLayout extends StatelessWidget {
final Widget mobile;
final Widget? tablet;
final Widget desktop;
const ResponsiveLayout({
super.key,
required this.mobile,
this.tablet,
required this.desktop,
});
@override
Widget build(BuildContext context) {
return LayoutBuilder(
builder: (context, constraints) {
if (constraints.maxWidth >= 1100) return desktop;
if (constraints.maxWidth >= 650) return tablet ?? mobile;
return mobile;
},
);
}
}
```
## Custom Hooks (flutter_hooks)
```dart
import 'package:flutter_hooks/flutter_hooks.dart';
class CounterWidget extends HookWidget {
@override
Widget build(BuildContext context) {
final counter = useState(0);
final controller = useTextEditingController();
useEffect(() {
// Setup
return () {
// Cleanup
};
}, []);
return Column(
children: [
Text('Count: ${counter.value}'),
ElevatedButton(
onPressed: () => counter.value++,
child: const Text('Increment'),
),
],
);
}
}
```
## Sliver Patterns
```dart
CustomScrollView(
slivers: [
SliverAppBar(
expandedHeight: 200,
pinned: true,
flexibleSpace: FlexibleSpaceBar(
title: const Text('Title'),
background: Image.network(imageUrl, fit: BoxFit.cover),
),
),
SliverList(
delegate: SliverChildBuilderDelegate(
(context, index) => ListTile(title: Text('Item $index')),
childCount: 100,
),
),
],
)
```
## Key Optimization Patterns
| Pattern | Implementation |
|---------|----------------|
| **const widgets** | Add `const` to static widgets |
| **keys** | Use `Key` for list items |
| **select** | `ref.watch(provider.select(...))` |
| **RepaintBoundary** | Isolate expensive repaints |
| **ListView.builder** | Lazy loading for lists |
| **const constructors** | Always use when possible |

42
skills/frontend-design/SKILL.md
View File

@ -1,42 +0,0 @@
---
name: frontend-design
description: "Anthropic 官方前端设计技能。创建独特的生产级前端界面，避免 AI 同质化美学。禁止 Inter/Roboto/Arial 等通用字体，强制大胆视觉选择。当用户要求构建 web 组件、页面或应用时使用。触发词: 前端设计、独特UI、anti-slop、distinctive design、bold aesthetics、creative frontend。"
maturity: stable
---
This skill guides creation of distinctive, production-grade frontend interfaces that avoid generic "AI slop" aesthetics. Implement real working code with exceptional attention to aesthetic details and creative choices.
The user provides frontend requirements: a component, page, application, or interface to build. They may include context about the purpose, audience, or technical constraints.
## Design Thinking
Before coding, understand the context and commit to a BOLD aesthetic direction:
- **Purpose**: What problem does this interface solve? Who uses it?
- **Tone**: Pick an extreme: brutally minimal, maximalist chaos, retro-futuristic, organic/natural, luxury/refined, playful/toy-like, editorial/magazine, brutalist/raw, art deco/geometric, soft/pastel, industrial/utilitarian, etc. There are so many flavors to choose from. Use these for inspiration but design one that is true to the aesthetic direction.
- **Constraints**: Technical requirements (framework, performance, accessibility).
- **Differentiation**: What makes this UNFORGETTABLE? What's the one thing someone will remember?
**CRITICAL**: Choose a clear conceptual direction and execute it with precision. Bold maximalism and refined minimalism both work - the key is intentionality, not intensity.
Then implement working code (HTML/CSS/JS, React, Vue, etc.) that is:
- Production-grade and functional
- Visually striking and memorable
- Cohesive with a clear aesthetic point-of-view
- Meticulously refined in every detail
## Frontend Aesthetics Guidelines
Focus on:
- **Typography**: Choose fonts that are beautiful, unique, and interesting. Avoid generic fonts like Arial and Inter; opt instead for distinctive choices that elevate the frontend's aesthetics; unexpected, characterful font choices. Pair a distinctive display font with a refined body font.
- **Color & Theme**: Commit to a cohesive aesthetic. Use CSS variables for consistency. Dominant colors with sharp accents outperform timid, evenly-distributed palettes.
- **Motion**: Use animations for effects and micro-interactions. Prioritize CSS-only solutions for HTML. Use Motion library for React when available. Focus on high-impact moments: one well-orchestrated page load with staggered reveals (animation-delay) creates more delight than scattered micro-interactions. Use scroll-triggering and hover states that surprise.
- **Spatial Composition**: Unexpected layouts. Asymmetry. Overlap. Diagonal flow. Grid-breaking elements. Generous negative space OR controlled density.
- **Backgrounds & Visual Details**: Create atmosphere and depth rather than defaulting to solid colors. Add contextual effects and textures that match the overall aesthetic. Apply creative forms like gradient meshes, noise textures, geometric patterns, layered transparencies, dramatic shadows, decorative borders, custom cursors, and grain overlays.
NEVER use generic AI-generated aesthetics like overused font families (Inter, Roboto, Arial, system fonts), cliched color schemes (particularly purple gradients on white backgrounds), predictable layouts and component patterns, and cookie-cutter design that lacks context-specific character.
Interpret creatively and make unexpected choices that feel genuinely designed for the context. No design should be the same. Vary between light and dark themes, different fonts, different aesthetics. NEVER converge on common choices (Space Grotesk, for example) across generations.
**IMPORTANT**: Match implementation complexity to the aesthetic vision. Maximalist designs need elaborate code with extensive animations and effects. Minimalist or refined designs need restraint, precision, and careful attention to spacing, typography, and subtle details. Elegance comes from executing the vision well.
Remember: Claude is capable of extraordinary creative work. Don't hold back, show what can truly be created when thinking outside the box and committing fully to a distinctive vision.

80
skills/golang-pro/SKILL.md
View File

@ -1,80 +0,0 @@
---
name: golang-pro
description: >
Go 语言深度专家。当用户需要 Go 1.21+ 并发编程、goroutine/channel、gRPC 服务、微服务架构、Go 性能优化、Go 模块管理，或说 "Go语言"、"Golang"、"goroutine" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, Bash
maturity: imported
last-reviewed: 2026-03-03
composable: true
enhances: [backend-builder, cloud-native-expert]
---
# Golang Pro
Senior Go developer with deep expertise in Go 1.21+, concurrent programming, and cloud-native microservices. Specializes in idiomatic patterns, performance optimization, and production-grade systems.
## Role Definition
You are a senior Go engineer with 8+ years of systems programming experience. You specialize in Go 1.21+ with generics, concurrent patterns, gRPC microservices, and cloud-native applications. You build efficient, type-safe systems following Go proverbs.
## When to Use This Skill
- Building concurrent Go applications with goroutines and channels
- Implementing microservices with gRPC or REST APIs
- Creating CLI tools and system utilities
- Optimizing Go code for performance and memory efficiency
- Designing interfaces and using Go generics
- Setting up testing with table-driven tests and benchmarks
## Core Workflow
1. **Analyze architecture** - Review module structure, interfaces, concurrency patterns
2. **Design interfaces** - Create small, focused interfaces with composition
3. **Implement** - Write idiomatic Go with proper error handling and context propagation
4. **Optimize** - Profile with pprof, write benchmarks, eliminate allocations
5. **Test** - Table-driven tests, race detector, fuzzing, 80%+ coverage
## Reference Guide
Load detailed guidance based on context:
| Topic | Reference | Load When |
|-------|-----------|-----------|
| Concurrency | `references/concurrency.md` | Goroutines, channels, select, sync primitives |
| Interfaces | `references/interfaces.md` | Interface design, io.Reader/Writer, composition |
| Generics | `references/generics.md` | Type parameters, constraints, generic patterns |
| Testing | `references/testing.md` | Table-driven tests, benchmarks, fuzzing |
| Project Structure | `references/project-structure.md` | Module layout, internal packages, go.mod |
## Constraints
### MUST DO
- Use gofmt and golangci-lint on all code
- Add context.Context to all blocking operations
- Handle all errors explicitly (no naked returns)
- Write table-driven tests with subtests
- Document all exported functions, types, and packages
- Use `X | Y` union constraints for generics (Go 1.18+)
- Propagate errors with fmt.Errorf("%w", err)
- Run race detector on tests (-race flag)
### MUST NOT DO
- Ignore errors (avoid _ assignment without justification)
- Use panic for normal error handling
- Create goroutines without clear lifecycle management
- Skip context cancellation handling
- Use reflection without performance justification
- Mix sync and async patterns carelessly
- Hardcode configuration (use functional options or env vars)
## Output Templates
When implementing Go features, provide:
1. Interface definitions (contracts first)
2. Implementation files with proper package structure
3. Test file with table-driven tests
4. Brief explanation of concurrency patterns used
## Knowledge Reference
Go 1.21+, goroutines, channels, select, sync package, generics, type parameters, constraints, io.Reader/Writer, gRPC, context, error wrapping, pprof profiling, benchmarks, table-driven tests, fuzzing, go.mod, internal packages, functional options

329
skills/golang-pro/references/concurrency.md
View File

@ -1,329 +0,0 @@
# Concurrency Patterns
## Goroutine Lifecycle Management
```go
package main
import (
"context"
"fmt"
"sync"
"time"
)
// Worker pool with bounded concurrency
type WorkerPool struct {
workers int
tasks chan func()
wg sync.WaitGroup
}
func NewWorkerPool(workers int) *WorkerPool {
wp := &WorkerPool{
workers: workers,
tasks: make(chan func(), workers*2), // Buffered channel
}
wp.start()
return wp
}
func (wp *WorkerPool) start() {
for i := 0; i < wp.workers; i++ {
wp.wg.Add(1)
go func() {
defer wp.wg.Done()
for task := range wp.tasks {
task()
}
}()
}
}
func (wp *WorkerPool) Submit(task func()) {
wp.tasks <- task
}
func (wp *WorkerPool) Shutdown() {
close(wp.tasks)
wp.wg.Wait()
}
```
## Channel Patterns
```go
// Generator pattern
func generateNumbers(ctx context.Context, max int) <-chan int {
out := make(chan int)
go func() {
defer close(out)
for i := 0; i < max; i++ {
select {
case out <- i:
case <-ctx.Done():
return
}
}
}()
return out
}
// Fan-out, fan-in pattern
func fanOut(ctx context.Context, input <-chan int, workers int) []<-chan int {
channels := make([]<-chan int, workers)
for i := 0; i < workers; i++ {
channels[i] = process(ctx, input)
}
return channels
}
func process(ctx context.Context, input <-chan int) <-chan int {
out := make(chan int)
go func() {
defer close(out)
for val := range input {
select {
case out <- val * 2:
case <-ctx.Done():
return
}
}
}()
return out
}
func fanIn(ctx context.Context, channels ...<-chan int) <-chan int {
out := make(chan int)
var wg sync.WaitGroup
for _, ch := range channels {
wg.Add(1)
go func(c <-chan int) {
defer wg.Done()
for val := range c {
select {
case out <- val:
case <-ctx.Done():
return
}
}
}(ch)
}
go func() {
wg.Wait()
close(out)
}()
return out
}
```
## Select Statement Patterns
```go
// Timeout pattern
func fetchWithTimeout(ctx context.Context, url string) (string, error) {
result := make(chan string, 1)
errCh := make(chan error, 1)
go func() {
// Simulate network call
time.Sleep(100 * time.Millisecond)
result <- "data from " + url
}()
select {
case res := <-result:
return res, nil
case err := <-errCh:
return "", err
case <-time.After(50 * time.Millisecond):
return "", fmt.Errorf("timeout")
case <-ctx.Done():
return "", ctx.Err()
}
}
// Done channel pattern for graceful shutdown
type Server struct {
done chan struct{}
}
func (s *Server) Shutdown() {
close(s.done)
}
func (s *Server) Run(ctx context.Context) {
ticker := time.NewTicker(1 * time.Second)
defer ticker.Stop()
for {
select {
case <-ticker.C:
fmt.Println("tick")
case <-s.done:
fmt.Println("shutting down")
return
case <-ctx.Done():
fmt.Println("context cancelled")
return
}
}
}
```
## Sync Primitives
```go
import "sync"
// Mutex for protecting shared state
type Counter struct {
mu sync.Mutex
count int
}
func (c *Counter) Increment() {
c.mu.Lock()
defer c.mu.Unlock()
c.count++
}
func (c *Counter) Value() int {
c.mu.Lock()
defer c.mu.Unlock()
return c.count
}
// RWMutex for read-heavy workloads
type Cache struct {
mu sync.RWMutex
items map[string]string
}
func (c *Cache) Get(key string) (string, bool) {
c.mu.RLock()
defer c.mu.RUnlock()
val, ok := c.items[key]
return val, ok
}
func (c *Cache) Set(key, value string) {
c.mu.Lock()
defer c.mu.Unlock()
c.items[key] = value
}
// sync.Once for initialization
type Service struct {
once sync.Once
config *Config
}
func (s *Service) getConfig() *Config {
s.once.Do(func() {
s.config = loadConfig() // Only called once
})
return s.config
}
```
## Rate Limiting and Backpressure
```go
import "golang.org/x/time/rate"
// Token bucket rate limiter
type RateLimiter struct {
limiter *rate.Limiter
}
func NewRateLimiter(rps int) *RateLimiter {
return &RateLimiter{
limiter: rate.NewLimiter(rate.Limit(rps), rps),
}
}
func (rl *RateLimiter) Process(ctx context.Context, item string) error {
if err := rl.limiter.Wait(ctx); err != nil {
return err
}
// Process item
return nil
}
// Semaphore pattern for limiting concurrency
type Semaphore struct {
slots chan struct{}
}
func NewSemaphore(n int) *Semaphore {
return &Semaphore{
slots: make(chan struct{}, n),
}
}
func (s *Semaphore) Acquire() {
s.slots <- struct{}{}
}
func (s *Semaphore) Release() {
<-s.slots
}
func (s *Semaphore) Do(fn func()) {
s.Acquire()
defer s.Release()
fn()
}
```
## Pipeline Pattern
```go
// Stage-based processing pipeline
func pipeline(ctx context.Context, input <-chan int) <-chan int {
// Stage 1: Square numbers
stage1 := make(chan int)
go func() {
defer close(stage1)
for num := range input {
select {
case stage1 <- num * num:
case <-ctx.Done():
return
}
}
}()
// Stage 2: Filter even numbers
stage2 := make(chan int)
go func() {
defer close(stage2)
for num := range stage1 {
if num%2 == 0 {
select {
case stage2 <- num:
case <-ctx.Done():
return
}
}
}
}()
return stage2
}
```
## Quick Reference
| Pattern | Use Case | Key Points |
|---------|----------|------------|
| Worker Pool | Bounded concurrency | Limit goroutines, reuse workers |
| Fan-out/Fan-in | Parallel processing | Distribute work, merge results |
| Pipeline | Stream processing | Chain transformations |
| Rate Limiter | API throttling | Control request rate |
| Semaphore | Resource limits | Cap concurrent operations |
| Done Channel | Graceful shutdown | Signal completion |

442
skills/golang-pro/references/generics.md
View File

@ -1,442 +0,0 @@
# Generics and Type Parameters
## Basic Type Parameters
```go
package main
// Generic function with type parameter
func Max[T constraints.Ordered](a, b T) T {
if a > b {
return a
}
return b
}
// Multiple type parameters
func Map[T, U any](slice []T, fn func(T) U) []U {
result := make([]U, len(slice))
for i, v := range slice {
result[i] = fn(v)
}
return result
}
// Usage
func main() {
maxInt := Max(10, 20) // T = int
maxFloat := Max(3.14, 2.71) // T = float64
maxString := Max("abc", "xyz") // T = string
nums := []int{1, 2, 3}
doubled := Map(nums, func(n int) int { return n * 2 })
strings := Map(nums, func(n int) string { return fmt.Sprintf("%d", n) })
}
```
## Type Constraints
```go
import "constraints"
// Built-in constraints
type Number interface {
constraints.Integer | constraints.Float
}
func Sum[T Number](numbers []T) T {
var total T
for _, n := range numbers {
total += n
}
return total
}
// Custom constraints with methods
type Stringer interface {
String() string
}
func PrintAll[T Stringer](items []T) {
for _, item := range items {
fmt.Println(item.String())
}
}
// Approximate constraint using ~
type Integer interface {
~int | ~int8 | ~int16 | ~int32 | ~int64
}
type MyInt int
func Double[T Integer](n T) T {
return n * 2
}
// Works with both int and MyInt
func main() {
fmt.Println(Double(5)) // int
fmt.Println(Double(MyInt(5))) // MyInt
}
```
## Generic Data Structures
```go
// Generic Stack
type Stack[T any] struct {
items []T
}
func NewStack[T any]() *Stack[T] {
return &Stack[T]{
items: make([]T, 0),
}
}
func (s *Stack[T]) Push(item T) {
s.items = append(s.items, item)
}
func (s *Stack[T]) Pop() (T, bool) {
if len(s.items) == 0 {
var zero T
return zero, false
}
item := s.items[len(s.items)-1]
s.items = s.items[:len(s.items)-1]
return item, true
}
func (s *Stack[T]) IsEmpty() bool {
return len(s.items) == 0
}
// Usage
intStack := NewStack[int]()
intStack.Push(1)
intStack.Push(2)
stringStack := NewStack[string]()
stringStack.Push("hello")
stringStack.Push("world")
```
## Generic Map Operations
```go
// Filter with generics
func Filter[T any](slice []T, predicate func(T) bool) []T {
result := make([]T, 0, len(slice))
for _, v := range slice {
if predicate(v) {
result = append(result, v)
}
}
return result
}
// Reduce/Fold
func Reduce[T, U any](slice []T, initial U, fn func(U, T) U) U {
acc := initial
for _, v := range slice {
acc = fn(acc, v)
}
return acc
}
// Keys from map
func Keys[K comparable, V any](m map[K]V) []K {
keys := make([]K, 0, len(m))
for k := range m {
keys = append(keys, k)
}
return keys
}
// Values from map
func Values[K comparable, V any](m map[K]V) []V {
values := make([]V, 0, len(m))
for _, v := range m {
values = append(values, v)
}
return values
}
// Usage
numbers := []int{1, 2, 3, 4, 5, 6}
evens := Filter(numbers, func(n int) bool { return n%2 == 0 })
sum := Reduce(numbers, 0, func(acc, n int) int { return acc + n })
m := map[string]int{"a": 1, "b": 2}
keys := Keys(m) // []string{"a", "b"}
values := Values(m) // []int{1, 2}
```
## Generic Pairs and Tuples
```go
// Generic Pair
type Pair[T, U any] struct {
First T
Second U
}
func NewPair[T, U any](first T, second U) Pair[T, U] {
return Pair[T, U]{First: first, Second: second}
}
func (p Pair[T, U]) Swap() Pair[U, T] {
return Pair[U, T]{First: p.Second, Second: p.First}
}
// Usage
pair := NewPair("name", 42)
swapped := pair.Swap() // Pair[int, string]
// Generic Result type (like Rust's Result<T, E>)
type Result[T any] struct {
value T
err error
}
func Ok[T any](value T) Result[T] {
return Result[T]{value: value}
}
func Err[T any](err error) Result[T] {
return Result[T]{err: err}
}
func (r Result[T]) IsOk() bool {
return r.err == nil
}
func (r Result[T]) Unwrap() (T, error) {
return r.value, r.err
}
func (r Result[T]) UnwrapOr(defaultValue T) T {
if r.err != nil {
return defaultValue
}
return r.value
}
```
## Comparable Constraint
```go
// Find using comparable
func Find[T comparable](slice []T, target T) (int, bool) {
for i, v := range slice {
if v == target {
return i, true
}
}
return -1, false
}
// Contains
func Contains[T comparable](slice []T, target T) bool {
_, found := Find(slice, target)
return found
}
// Unique elements
func Unique[T comparable](slice []T) []T {
seen := make(map[T]struct{})
result := make([]T, 0, len(slice))
for _, v := range slice {
if _, exists := seen[v]; !exists {
seen[v] = struct{}{}
result = append(result, v)
}
}
return result
}
// Usage
nums := []int{1, 2, 2, 3, 3, 4}
unique := Unique(nums) // []int{1, 2, 3, 4}
idx, found := Find([]string{"a", "b", "c"}, "b") // 1, true
```
## Generic Interfaces
```go
// Generic interface
type Container[T any] interface {
Add(item T)
Remove() (T, bool)
Size() int
}
// Implementation
type Queue[T any] struct {
items []T
}
func (q *Queue[T]) Add(item T) {
q.items = append(q.items, item)
}
func (q *Queue[T]) Remove() (T, bool) {
if len(q.items) == 0 {
var zero T
return zero, false
}
item := q.items[0]
q.items = q.items[1:]
return item, true
}
func (q *Queue[T]) Size() int {
return len(q.items)
}
// Function accepting generic interface
func ProcessContainer[T any](c Container[T], item T) {
c.Add(item)
fmt.Printf("Container size: %d\n", c.Size())
}
```
## Type Inference
```go
// Type inference works in most cases
func Identity[T any](x T) T {
return x
}
// No need to specify type
result := Identity(42) // T inferred as int
str := Identity("hello") // T inferred as string
// Type inference with constraints
func Min[T constraints.Ordered](a, b T) T {
if a < b {
return a
}
return b
}
// Inferred from arguments
minVal := Min(10, 20) // T = int
minFloat := Min(1.5, 2.5) // T = float64
// Explicit type when needed
result := Map[int, string]([]int{1, 2}, func(n int) string {
return fmt.Sprintf("%d", n)
})
```
## Generic Channels
```go
// Generic channel operations
func Merge[T any](channels ...<-chan T) <-chan T {
out := make(chan T)
var wg sync.WaitGroup
for _, ch := range channels {
wg.Add(1)
go func(c <-chan T) {
defer wg.Done()
for v := range c {
out <- v
}
}(ch)
}
go func() {
wg.Wait()
close(out)
}()
return out
}
// Generic pipeline stage
func Stage[T, U any](in <-chan T, fn func(T) U) <-chan U {
out := make(chan U)
go func() {
defer close(out)
for v := range in {
out <- fn(v)
}
}()
return out
}
// Usage
ch1 := make(chan int)
ch2 := make(chan int)
merged := Merge(ch1, ch2)
numbers := make(chan int)
doubled := Stage(numbers, func(n int) int { return n * 2 })
strings := Stage(doubled, func(n int) string { return fmt.Sprintf("%d", n) })
```
## Union Constraints
```go
// Union of types
type StringOrInt interface {
string | int
}
func Process[T StringOrInt](val T) string {
return fmt.Sprintf("%v", val)
}
// More complex unions
type Numeric interface {
int | int8 | int16 | int32 | int64 |
uint | uint8 | uint16 | uint32 | uint64 |
float32 | float64
}
func Abs[T Numeric](n T) T {
if n < 0 {
return -n
}
return n
}
// Union with methods
type Serializable interface {
string | []byte
}
func Serialize[T Serializable](data T) []byte {
switch v := any(data).(type) {
case string:
return []byte(v)
case []byte:
return v
default:
panic("unreachable")
}
}
```
## Quick Reference
| Feature | Syntax | Use Case |
|---------|--------|----------|
| Basic generic | `func F[T any]()` | Any type |
| Constraint | `func F[T Constraint]()` | Restricted types |
| Multiple params | `func F[T, U any]()` | Multiple type variables |
| Comparable | `func F[T comparable]()` | Types supporting == and != |
| Ordered | `func F[T constraints.Ordered]()` | Types supporting <, >, <=, >= |
| Union | `T interface{int \| string}` | Either type |
| Approximate | `~int` | Include type aliases |

432
skills/golang-pro/references/interfaces.md
View File

@ -1,432 +0,0 @@
# Interface Design and Composition
## Small, Focused Interfaces
```go
// Single-method interfaces (idiomatic Go)
type Reader interface {
Read(p []byte) (n int, err error)
}
type Writer interface {
Write(p []byte) (n int, err error)
}
type Closer interface {
Close() error
}
// Interface composition
type ReadCloser interface {
Reader
Closer
}
type WriteCloser interface {
Writer
Closer
}
type ReadWriteCloser interface {
Reader
Writer
Closer
}
```
## Accept Interfaces, Return Structs
```go
package storage
import "io"
// Storage is the concrete type (struct)
type Storage struct {
baseDir string
}
// NewStorage returns a concrete type
func NewStorage(baseDir string) *Storage {
return &Storage{baseDir: baseDir}
}
// SaveFile accepts an interface for flexibility
func (s *Storage) SaveFile(filename string, data io.Reader) error {
// Implementation can work with any Reader
// (file, network, buffer, etc.)
return nil
}
// Usage allows dependency injection
type Uploader interface {
SaveFile(filename string, data io.Reader) error
}
type Service struct {
uploader Uploader // Accept interface
}
// NewService accepts interface for testing flexibility
func NewService(uploader Uploader) *Service {
return &Service{uploader: uploader}
}
```
## io.Reader and io.Writer Patterns
```go
import (
"io"
"strings"
)
// Chain readers with io.MultiReader
func combineReaders() io.Reader {
r1 := strings.NewReader("Hello ")
r2 := strings.NewReader("World")
return io.MultiReader(r1, r2)
}
// Tee reader for duplicating reads
func duplicateRead(r io.Reader, w io.Writer) io.Reader {
return io.TeeReader(r, w) // Writes to w while reading from r
}
// Limit reader to prevent reading too much
func limitedRead(r io.Reader, n int64) io.Reader {
return io.LimitReader(r, n)
}
// Custom Reader implementation
type UppercaseReader struct {
src io.Reader
}
func (u *UppercaseReader) Read(p []byte) (n int, err error) {
n, err = u.src.Read(p)
for i := 0; i < n; i++ {
if p[i] >= 'a' && p[i] <= 'z' {
p[i] = p[i] - 32
}
}
return n, err
}
// Custom Writer implementation
type CountingWriter struct {
w io.Writer
count int64
}
func (cw *CountingWriter) Write(p []byte) (n int, err error) {
n, err = cw.w.Write(p)
cw.count += int64(n)
return n, err
}
func (cw *CountingWriter) BytesWritten() int64 {
return cw.count
}
```
## Embedding for Composition
```go
import "sync"
// Embed to extend behavior
type SafeCounter struct {
mu sync.Mutex
m map[string]int
}
func (sc *SafeCounter) Inc(key string) {
sc.mu.Lock()
defer sc.mu.Unlock()
sc.m[key]++
}
// Embed interface to add default behavior
type Logger interface {
Log(msg string)
}
type NoOpLogger struct{}
func (NoOpLogger) Log(msg string) {}
type Service struct {
Logger // Embedded interface (default implementation can be provided)
}
func NewService(logger Logger) *Service {
if logger == nil {
logger = NoOpLogger{} // Provide default
}
return &Service{Logger: logger}
}
// Now Service.Log() is available
```
## Interface Satisfaction Verification
```go
import "io"
// Compile-time interface verification
var _ io.Reader = (*MyReader)(nil)
var _ io.Writer = (*MyWriter)(nil)
var _ io.Closer = (*MyCloser)(nil)
type MyReader struct{}
func (m *MyReader) Read(p []byte) (n int, err error) {
return 0, nil
}
type MyWriter struct{}
func (m *MyWriter) Write(p []byte) (n int, err error) {
return len(p), nil
}
type MyCloser struct{}
func (m *MyCloser) Close() error {
return nil
}
```
## Functional Options Pattern
```go
package server
import "time"
type Server struct {
host string
port int
timeout time.Duration
maxConns int
enableLogger bool
}
// Option is a functional option for configuring Server
type Option func(*Server)
func WithHost(host string) Option {
return func(s *Server) {
s.host = host
}
}
func WithPort(port int) Option {
return func(s *Server) {
s.port = port
}
}
func WithTimeout(timeout time.Duration) Option {
return func(s *Server) {
s.timeout = timeout
}
}
func WithMaxConnections(max int) Option {
return func(s *Server) {
s.maxConns = max
}
}
func WithLogger(enabled bool) Option {
return func(s *Server) {
s.enableLogger = enabled
}
}
// NewServer creates a server with functional options
func NewServer(opts ...Option) *Server {
// Defaults
s := &Server{
host: "localhost",
port: 8080,
timeout: 30 * time.Second,
maxConns: 100,
}
// Apply options
for _, opt := range opts {
opt(s)
}
return s
}
// Usage:
// server := NewServer(
// WithHost("0.0.0.0"),
// WithPort(9000),
// WithTimeout(60 * time.Second),
// WithLogger(true),
// )
```
## Interface Segregation
```go
// Bad: Fat interface
type BadRepository interface {
Create(item Item) error
Read(id string) (Item, error)
Update(item Item) error
Delete(id string) error
List() ([]Item, error)
Search(query string) ([]Item, error)
Count() (int, error)
}
// Good: Segregated interfaces
type Creator interface {
Create(item Item) error
}
type Reader interface {
Read(id string) (Item, error)
}
type Updater interface {
Update(item Item) error
}
type Deleter interface {
Delete(id string) error
}
type Lister interface {
List() ([]Item, error)
}
// Compose only what you need
type ReadWriter interface {
Reader
Creator
}
type FullRepository interface {
Creator
Reader
Updater
Deleter
Lister
}
```
## Type Assertions and Type Switches
```go
import "fmt"
// Safe type assertion
func processValue(v interface{}) {
// Two-value assertion (safe)
if str, ok := v.(string); ok {
fmt.Println("String:", str)
return
}
// Type switch
switch val := v.(type) {
case int:
fmt.Println("Int:", val)
case string:
fmt.Println("String:", val)
case bool:
fmt.Println("Bool:", val)
default:
fmt.Println("Unknown type")
}
}
// Check for optional interface methods
type Flusher interface {
Flush() error
}
func writeAndFlush(w io.Writer, data []byte) error {
if _, err := w.Write(data); err != nil {
return err
}
// Check if Writer also implements Flusher
if flusher, ok := w.(Flusher); ok {
return flusher.Flush()
}
return nil
}
```
## Dependency Injection via Interfaces
```go
package app
import "context"
// Define interfaces for dependencies
type UserRepository interface {
GetUser(ctx context.Context, id string) (*User, error)
SaveUser(ctx context.Context, user *User) error
}
type EmailSender interface {
SendEmail(ctx context.Context, to, subject, body string) error
}
// Service depends on interfaces
type UserService struct {
repo UserRepository
mailer EmailSender
}
func NewUserService(repo UserRepository, mailer EmailSender) *UserService {
return &UserService{
repo: repo,
mailer: mailer,
}
}
func (s *UserService) RegisterUser(ctx context.Context, email string) error {
user := &User{Email: email}
if err := s.repo.SaveUser(ctx, user); err != nil {
return err
}
return s.mailer.SendEmail(ctx, email, "Welcome", "Thanks for registering!")
}
// Easy to mock in tests
type MockUserRepository struct{}
func (m *MockUserRepository) GetUser(ctx context.Context, id string) (*User, error) {
return &User{ID: id}, nil
}
func (m *MockUserRepository) SaveUser(ctx context.Context, user *User) error {
return nil
}
```
## Quick Reference
| Pattern | Use Case | Key Principle |
|---------|----------|---------------|
| Small interfaces | Flexibility | Single-method interfaces |
| Accept interfaces | Testability | Depend on abstractions |
| Return structs | Clarity | Concrete return types |
| io.Reader/Writer | I/O operations | Standard library integration |
| Embedding | Composition | Extend behavior without inheritance |
| Functional options | Configuration | Flexible constructors |
| Type assertions | Runtime checks | Safe downcasting |

477
skills/golang-pro/references/project-structure.md
View File

@ -1,477 +0,0 @@
# Project Structure and Module Management
## Standard Project Layout
```
myproject/
├── cmd/ # Main applications
│ ├── server/
│ │ └── main.go # Entry point for server
│ └── cli/
│ └── main.go # Entry point for CLI tool
├── internal/ # Private application code
│ ├── api/ # API handlers
│ ├── service/ # Business logic
│ └── repository/ # Data access layer
├── pkg/ # Public library code
│ └── models/ # Shared models
├── api/ # API definitions
│ ├── openapi.yaml # OpenAPI spec
│ └── proto/ # Protocol buffers
├── web/ # Web assets
│ ├── static/
│ └── templates/
├── scripts/ # Build and install scripts
├── configs/ # Configuration files
├── deployments/ # Docker, K8s configs
├── test/ # Additional test data
├── docs/ # Documentation
├── go.mod # Module definition
├── go.sum # Dependency checksums
├── Makefile # Build automation
└── README.md
```
## go.mod Basics
```go
// Initialize module
// go mod init github.com/user/project
module github.com/user/myproject
go 1.21
require (
github.com/gin-gonic/gin v1.9.1
github.com/lib/pq v1.10.9
go.uber.org/zap v1.26.0
)
require (
// Indirect dependencies (automatically managed)
github.com/bytedance/sonic v1.9.1 // indirect
github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311 // indirect
)
// Replace directive for local development
replace github.com/user/mylib => ../mylib
// Retract directive to mark bad versions
retract v1.0.1 // Contains critical bug
```
## Module Commands
```bash
# Initialize module
go mod init github.com/user/project
# Add missing dependencies
go mod tidy
# Download dependencies
go mod download
# Verify dependencies
go mod verify
# Show module graph
go mod graph
# Show why package is needed
go mod why github.com/user/package
# Vendor dependencies (copy to vendor/)
go mod vendor
# Update dependency
go get -u github.com/user/package
# Update to specific version
go get github.com/user/package@v1.2.3
# Update all dependencies
go get -u ./...
# Remove unused dependencies
go mod tidy
```
## Internal Packages
```go
// internal/ packages can only be imported by code in the parent tree
myproject/
├── internal/
│ ├── auth/ # Can only be imported by myproject
│ │ └── jwt.go
│ └── database/
│ └── postgres.go
└── pkg/
└── models/ # Can be imported by anyone
└── user.go
// This works (same project):
import "github.com/user/myproject/internal/auth"
// This fails (different project):
import "github.com/other/project/internal/auth" // Error!
// Internal subdirectories
myproject/
└── api/
└── internal/ # Can only be imported by code in api/
└── helpers.go
```
## Package Organization
```go
// user/user.go - Domain package
package user
import (
"context"
"time"
)
// User represents a user entity
type User struct {
ID string
Email string
CreatedAt time.Time
}
// Repository defines data access interface
type Repository interface {
Create(ctx context.Context, user *User) error
GetByID(ctx context.Context, id string) (*User, error)
Update(ctx context.Context, user *User) error
Delete(ctx context.Context, id string) error
}
// Service handles business logic
type Service struct {
repo Repository
}
// NewService creates a new user service
func NewService(repo Repository) *Service {
return &Service{repo: repo}
}
func (s *Service) RegisterUser(ctx context.Context, email string) (*User, error) {
user := &User{
ID: generateID(),
Email: email,
CreatedAt: time.Now(),
}
return user, s.repo.Create(ctx, user)
}
```
## Multi-Module Repository (Monorepo)
```
monorepo/
├── go.work # Workspace file
├── services/
│ ├── api/
│ │ ├── go.mod
│ │ └── main.go
│ └── worker/
│ ├── go.mod
│ └── main.go
└── shared/
└── models/
├── go.mod
└── user.go
// go.work
go 1.21
use (
./services/api
./services/worker
./shared/models
)
// Commands:
// go work init ./services/api ./services/worker
// go work use ./shared/models
// go work sync
```
## Build Tags and Constraints
```go
// +build integration
// integration_test.go
package myapp
import "testing"
func TestIntegration(t *testing.T) {
// Integration test code
}
// Build: go test -tags=integration
// File-level build constraints (Go 1.17+)
//go:build linux && amd64
package myapp
// Multiple constraints
//go:build linux || darwin
//go:build amd64
// Negation
//go:build !windows
// Common tags:
// linux, darwin, windows, freebsd
// amd64, arm64, 386, arm
// cgo, !cgo
```
## Makefile Example
```makefile
# Makefile
.PHONY: build test lint clean run
# Variables
BINARY_NAME=myapp
BUILD_DIR=bin
GO=go
GOFLAGS=-v
# Build the application
build:
$(GO) build $(GOFLAGS) -o $(BUILD_DIR)/$(BINARY_NAME) ./cmd/server
# Run tests
test:
$(GO) test -v -race -coverprofile=coverage.out ./...
# Run tests with coverage report
test-coverage: test
$(GO) tool cover -html=coverage.out
# Run linters
lint:
golangci-lint run ./...
# Format code
fmt:
$(GO) fmt ./...
goimports -w .
# Run the application
run:
$(GO) run ./cmd/server
# Clean build artifacts
clean:
rm -rf $(BUILD_DIR)
rm -f coverage.out
# Install dependencies
deps:
$(GO) mod download
$(GO) mod tidy
# Build for multiple platforms
build-all:
GOOS=linux GOARCH=amd64 $(GO) build -o $(BUILD_DIR)/$(BINARY_NAME)-linux-amd64 ./cmd/server
GOOS=darwin GOARCH=amd64 $(GO) build -o $(BUILD_DIR)/$(BINARY_NAME)-darwin-amd64 ./cmd/server
GOOS=windows GOARCH=amd64 $(GO) build -o $(BUILD_DIR)/$(BINARY_NAME)-windows-amd64.exe ./cmd/server
# Run with race detector
run-race:
$(GO) run -race ./cmd/server
# Generate code
generate:
$(GO) generate ./...
# Docker build
docker-build:
docker build -t $(BINARY_NAME):latest .
# Help
help:
@echo "Available targets:"
@echo " build - Build the application"
@echo " test - Run tests"
@echo " test-coverage - Run tests with coverage report"
@echo " lint - Run linters"
@echo " fmt - Format code"
@echo " run - Run the application"
@echo " clean - Clean build artifacts"
@echo " deps - Install dependencies"
```
## Dockerfile Multi-Stage Build
```dockerfile
# Build stage
FROM golang:1.21-alpine AS builder
WORKDIR /app
# Copy go mod files
COPY go.mod go.sum ./
RUN go mod download
# Copy source code
COPY . .
# Build binary
RUN CGO_ENABLED=0 GOOS=linux go build -a -installsuffix cgo -o server ./cmd/server
# Final stage
FROM alpine:latest
RUN apk --no-cache add ca-certificates
WORKDIR /root/
# Copy binary from builder
COPY --from=builder /app/server .
# Copy config files if needed
COPY --from=builder /app/configs ./configs
EXPOSE 8080
CMD ["./server"]
```
## Version Information
```go
// version/version.go
package version
import "runtime"
var (
// Set via ldflags during build
Version = "dev"
GitCommit = "none"
BuildTime = "unknown"
)
// Info returns version information
func Info() map[string]string {
return map[string]string{
"version": Version,
"git_commit": GitCommit,
"build_time": BuildTime,
"go_version": runtime.Version(),
"os": runtime.GOOS,
"arch": runtime.GOARCH,
}
}
// Build with version info:
// go build -ldflags "-X github.com/user/project/version.Version=1.0.0 \
// -X github.com/user/project/version.GitCommit=$(git rev-parse HEAD) \
// -X github.com/user/project/version.BuildTime=$(date -u +%Y-%m-%dT%H:%M:%SZ)"
```
## Go Generate
```go
// models/user.go
//go:generate mockgen -source=user.go -destination=../mocks/user_mock.go -package=mocks
package models
type UserRepository interface {
GetUser(id string) (*User, error)
SaveUser(user *User) error
}
// tools.go - Track tool dependencies
//go:build tools
package tools
import (
_ "github.com/golang/mock/mockgen"
_ "golang.org/x/tools/cmd/stringer"
)
// Install tools:
// go install github.com/golang/mock/mockgen@latest
// Run generate:
// go generate ./...
```
## Configuration Management
```go
// config/config.go
package config
import (
"os"
"time"
"github.com/kelseyhightower/envconfig"
)
type Config struct {
Server ServerConfig
Database DatabaseConfig
Redis RedisConfig
}
type ServerConfig struct {
Host string `envconfig:"SERVER_HOST" default:"0.0.0.0"`
Port int `envconfig:"SERVER_PORT" default:"8080"`
ReadTimeout time.Duration `envconfig:"SERVER_READ_TIMEOUT" default:"10s"`
WriteTimeout time.Duration `envconfig:"SERVER_WRITE_TIMEOUT" default:"10s"`
}
type DatabaseConfig struct {
URL string `envconfig:"DATABASE_URL" required:"true"`
MaxOpenConns int `envconfig:"DB_MAX_OPEN_CONNS" default:"25"`
MaxIdleConns int `envconfig:"DB_MAX_IDLE_CONNS" default:"5"`
}
type RedisConfig struct {
Addr string `envconfig:"REDIS_ADDR" default:"localhost:6379"`
Password string `envconfig:"REDIS_PASSWORD"`
DB int `envconfig:"REDIS_DB" default:"0"`
}
// Load loads configuration from environment
func Load() (*Config, error) {
var cfg Config
if err := envconfig.Process("", &cfg); err != nil {
return nil, err
}
return &cfg, nil
}
```
## Quick Reference
| Command | Description |
|---------|-------------|
| `go mod init` | Initialize module |
| `go mod tidy` | Add/remove dependencies |
| `go mod download` | Download dependencies |
| `go get package@version` | Add/update dependency |
| `go build -ldflags "-X ..."` | Set version info |
| `go generate ./...` | Run code generation |
| `GOOS=linux go build` | Cross-compile |
| `go work init` | Initialize workspace |

451
skills/golang-pro/references/testing.md
View File

@ -1,451 +0,0 @@
# Testing and Benchmarking
## Table-Driven Tests
```go
package math
import "testing"
func Add(a, b int) int {
return a + b
}
func TestAdd(t *testing.T) {
tests := []struct {
name string
a, b int
expected int
}{
{"positive numbers", 2, 3, 5},
{"negative numbers", -2, -3, -5},
{"mixed signs", -2, 3, 1},
{"zeros", 0, 0, 0},
{"large numbers", 1000000, 2000000, 3000000},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := Add(tt.a, tt.b)
if result != tt.expected {
t.Errorf("Add(%d, %d) = %d; want %d", tt.a, tt.b, result, tt.expected)
}
})
}
}
```
## Subtests and Parallel Execution
```go
func TestParallel(t *testing.T) {
tests := []struct {
name string
input string
want string
}{
{"lowercase", "hello", "HELLO"},
{"uppercase", "WORLD", "WORLD"},
{"mixed", "HeLLo", "HELLO"},
}
for _, tt := range tests {
tt := tt // Capture range variable for parallel tests
t.Run(tt.name, func(t *testing.T) {
t.Parallel() // Run subtests in parallel
result := strings.ToUpper(tt.input)
if result != tt.want {
t.Errorf("got %q, want %q", result, tt.want)
}
})
}
}
```
## Test Helpers and Setup/Teardown
```go
func TestWithSetup(t *testing.T) {
// Setup
db := setupTestDB(t)
defer cleanupTestDB(t, db)
tests := []struct {
name string
user User
}{
{"valid user", User{Name: "John", Email: "john@example.com"}},
{"empty name", User{Name: "", Email: "test@example.com"}},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
err := db.SaveUser(tt.user)
if err != nil {
t.Fatalf("SaveUser failed: %v", err)
}
})
}
}
// Helper function (doesn't show in stack trace)
func setupTestDB(t *testing.T) *DB {
t.Helper()
db, err := NewDB(":memory:")
if err != nil {
t.Fatalf("failed to create test DB: %v", err)
}
return db
}
func cleanupTestDB(t *testing.T, db *DB) {
t.Helper()
if err := db.Close(); err != nil {
t.Errorf("failed to close DB: %v", err)
}
}
```
## Mocking with Interfaces
```go
// Interface to mock
type EmailSender interface {
Send(to, subject, body string) error
}
// Mock implementation
type MockEmailSender struct {
SentEmails []Email
ShouldFail bool
}
type Email struct {
To, Subject, Body string
}
func (m *MockEmailSender) Send(to, subject, body string) error {
if m.ShouldFail {
return fmt.Errorf("failed to send email")
}
m.SentEmails = append(m.SentEmails, Email{to, subject, body})
return nil
}
// Test using mock
func TestUserService_Register(t *testing.T) {
mockSender := &MockEmailSender{}
service := NewUserService(mockSender)
err := service.Register("user@example.com")
if err != nil {
t.Fatalf("Register failed: %v", err)
}
if len(mockSender.SentEmails) != 1 {
t.Errorf("expected 1 email sent, got %d", len(mockSender.SentEmails))
}
email := mockSender.SentEmails[0]
if email.To != "user@example.com" {
t.Errorf("expected email to user@example.com, got %s", email.To)
}
}
```
## Benchmarking
```go
func BenchmarkAdd(b *testing.B) {
for i := 0; i < b.N; i++ {
Add(100, 200)
}
}
// Benchmark with subtests
func BenchmarkStringOperations(b *testing.B) {
benchmarks := []struct {
name string
input string
}{
{"short", "hello"},
{"medium", strings.Repeat("hello", 10)},
{"long", strings.Repeat("hello", 100)},
}
for _, bm := range benchmarks {
b.Run(bm.name, func(b *testing.B) {
for i := 0; i < b.N; i++ {
_ = strings.ToUpper(bm.input)
}
})
}
}
// Benchmark with setup
func BenchmarkMapOperations(b *testing.B) {
m := make(map[string]int)
for i := 0; i < 1000; i++ {
m[fmt.Sprintf("key%d", i)] = i
}
b.ResetTimer() // Don't count setup time
for i := 0; i < b.N; i++ {
_ = m["key500"]
}
}
// Parallel benchmark
func BenchmarkConcurrentAccess(b *testing.B) {
var counter int64
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
atomic.AddInt64(&counter, 1)
}
})
}
// Memory allocation benchmark
func BenchmarkAllocation(b *testing.B) {
b.ReportAllocs() // Report allocations
for i := 0; i < b.N; i++ {
s := make([]int, 1000)
_ = s
}
}
```
## Fuzzing (Go 1.18+)
```go
func FuzzReverse(f *testing.F) {
// Seed corpus
testcases := []string{"hello", "world", "123", ""}
for _, tc := range testcases {
f.Add(tc)
}
f.Fuzz(func(t *testing.T, input string) {
reversed := Reverse(input)
doubleReversed := Reverse(reversed)
if input != doubleReversed {
t.Errorf("Reverse(Reverse(%q)) = %q, want %q", input, doubleReversed, input)
}
})
}
// Fuzz with multiple parameters
func FuzzAdd(f *testing.F) {
f.Add(1, 2)
f.Add(0, 0)
f.Add(-1, 1)
f.Fuzz(func(t *testing.T, a, b int) {
result := Add(a, b)
// Properties that should always hold
if result < a && b >= 0 {
t.Errorf("Add(%d, %d) = %d; result should be >= a when b >= 0", a, b, result)
}
})
}
```
## Test Coverage
```go
// Run tests with coverage:
// go test -cover
// go test -coverprofile=coverage.out
// go tool cover -html=coverage.out
func TestCalculate(t *testing.T) {
tests := []struct {
name string
input int
expected int
}{
{"zero", 0, 0},
{"positive", 5, 25},
{"negative", -3, 9},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := Calculate(tt.input)
if result != tt.expected {
t.Errorf("Calculate(%d) = %d; want %d", tt.input, result, tt.expected)
}
})
}
}
```
## Race Detector
```go
// Run with: go test -race
func TestConcurrentAccess(t *testing.T) {
var counter int
var wg sync.WaitGroup
// This will fail with -race if not synchronized
for i := 0; i < 10; i++ {
wg.Add(1)
go func() {
defer wg.Done()
counter++ // Data race!
}()
}
wg.Wait()
}
// Fixed version with mutex
func TestConcurrentAccessSafe(t *testing.T) {
var counter int
var mu sync.Mutex
var wg sync.WaitGroup
for i := 0; i < 10; i++ {
wg.Add(1)
go func() {
defer wg.Done()
mu.Lock()
counter++
mu.Unlock()
}()
}
wg.Wait()
if counter != 10 {
t.Errorf("expected 10, got %d", counter)
}
}
```
## Golden Files
```go
import (
"os"
"path/filepath"
"testing"
)
func TestRenderHTML(t *testing.T) {
data := Data{Title: "Test", Content: "Hello"}
result := RenderHTML(data)
goldenFile := filepath.Join("testdata", "expected.html")
if *update {
// Update golden file: go test -update
os.WriteFile(goldenFile, []byte(result), 0644)
}
expected, err := os.ReadFile(goldenFile)
if err != nil {
t.Fatalf("failed to read golden file: %v", err)
}
if result != string(expected) {
t.Errorf("output doesn't match golden file\ngot:\n%s\nwant:\n%s", result, expected)
}
}
var update = flag.Bool("update", false, "update golden files")
```
## Integration Tests
```go
// integration_test.go
// +build integration
package myapp
import (
"testing"
"time"
)
func TestIntegration(t *testing.T) {
if testing.Short() {
t.Skip("skipping integration test in short mode")
}
// Long-running integration test
server := startTestServer(t)
defer server.Stop()
time.Sleep(100 * time.Millisecond) // Wait for server
client := NewClient(server.URL)
resp, err := client.Get("/health")
if err != nil {
t.Fatalf("health check failed: %v", err)
}
if resp.Status != "ok" {
t.Errorf("expected status ok, got %s", resp.Status)
}
}
// Run: go test -tags=integration
// Run short tests only: go test -short
```
## Testable Examples
```go
// Example tests that appear in godoc
func ExampleAdd() {
result := Add(2, 3)
fmt.Println(result)
// Output: 5
}
func ExampleAdd_negative() {
result := Add(-2, -3)
fmt.Println(result)
// Output: -5
}
// Unordered output
func ExampleKeys() {
m := map[string]int{"a": 1, "b": 2, "c": 3}
keys := Keys(m)
for _, k := range keys {
fmt.Println(k)
}
// Unordered output:
// a
// b
// c
}
```
## Quick Reference
| Command | Description |
|---------|-------------|
| `go test` | Run tests |
| `go test -v` | Verbose output |
| `go test -run TestName` | Run specific test |
| `go test -bench .` | Run benchmarks |
| `go test -cover` | Show coverage |
| `go test -race` | Run race detector |
| `go test -short` | Skip long tests |
| `go test -fuzz FuzzName` | Run fuzzing |
| `go test -cpuprofile cpu.prof` | CPU profiling |
| `go test -memprofile mem.prof` | Memory profiling |

82
skills/graphql-architect/SKILL.md
View File

@ -1,82 +0,0 @@
---
name: graphql-architect
description: >
GraphQL 架构专家。当用户需要 GraphQL Schema 设计、Resolver 实现、订阅 Subscription、Apollo/Relay 客户端、N+1 问题解决、DataLoader、GraphQL 缓存策略，或说 "GraphQL"、"Schema设计"、"Apollo" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, Bash
maturity: imported
last-reviewed: 2026-03-03
composable: true
enhances: [api-designer, frontend-expert]
---
# GraphQL Architect
Senior GraphQL architect specializing in schema design and distributed graph architectures with deep expertise in Apollo Federation 2.5+, GraphQL subscriptions, and performance optimization.
## Role Definition
You are a senior GraphQL architect with 10+ years of API design experience. You specialize in Apollo Federation, schema-first design, and building type-safe API graphs that scale across teams and services. You master resolvers, DataLoader patterns, and real-time subscriptions.
## When to Use This Skill
- Designing GraphQL schemas and type systems
- Implementing Apollo Federation architectures
- Building resolvers with DataLoader optimization
- Creating real-time GraphQL subscriptions
- Optimizing query complexity and performance
- Setting up authentication and authorization
## Core Workflow
1. **Domain Modeling** - Map business domains to GraphQL type system
2. **Design Schema** - Create types, interfaces, unions with federation directives
3. **Implement Resolvers** - Write efficient resolvers with DataLoader patterns
4. **Secure** - Add query complexity limits, depth limiting, field-level auth
5. **Optimize** - Performance tune with caching, persisted queries, monitoring
## Reference Guide
Load detailed guidance based on context:
| Topic | Reference | Load When |
|-------|-----------|-----------|
| Schema Design | `references/schema-design.md` | Types, interfaces, unions, enums, input types |
| Resolvers | `references/resolvers.md` | Resolver patterns, context, DataLoader, N+1 |
| Federation | `references/federation.md` | Apollo Federation, subgraphs, entities, directives |
| Subscriptions | `references/subscriptions.md` | Real-time updates, WebSocket, pub/sub patterns |
| Security | `references/security.md` | Query depth, complexity analysis, authentication |
| REST Migration | `references/migration-from-rest.md` | Migrating REST APIs to GraphQL |
## Constraints
### MUST DO
- Use schema-first design approach
- Implement proper nullable field patterns
- Use DataLoader for batching and caching
- Add query complexity analysis
- Document all types and fields
- Follow GraphQL naming conventions (camelCase)
- Use federation directives correctly
- Provide example queries for all operations
### MUST NOT DO
- Create N+1 query problems
- Skip query depth limiting
- Expose internal implementation details
- Use REST patterns in GraphQL
- Return null for non-nullable fields
- Skip error handling in resolvers
- Hardcode authorization logic
- Ignore schema validation
## Output Templates
When implementing GraphQL features, provide:
1. Schema definition (SDL with types and directives)
2. Resolver implementation (with DataLoader patterns)
3. Query/mutation/subscription examples
4. Brief explanation of design decisions
## Knowledge Reference
Apollo Server, Apollo Federation 2.5+, GraphQL SDL, DataLoader, GraphQL Subscriptions, WebSocket, Redis pub/sub, schema composition, query complexity, persisted queries, schema stitching, type generation

418
skills/graphql-architect/references/federation.md
View File

@ -1,418 +0,0 @@
# Apollo Federation
## Subgraph Setup
```typescript
// users-subgraph/schema.graphql
extend schema
@link(url: "https://specs.apollo.dev/federation/v2.5", import: ["@key", "@shareable"])
type User @key(fields: "id") {
id: ID!
email: String!
username: String!
createdAt: DateTime!
}
type Query {
user(id: ID!): User
users: [User!]!
}
// users-subgraph/resolvers.ts
import { ApolloServer } from '@apollo/server';
import { buildSubgraphSchema } from '@apollo/subgraph';
import { readFileSync } from 'fs';
const typeDefs = readFileSync('./schema.graphql', 'utf8');
const resolvers = {
User: {
__resolveReference: async (
reference: { id: string },
context: Context
): Promise<User> => {
return context.dataSources.users.findById(reference.id);
},
},
Query: {
user: async (parent, args: { id: string }, context: Context) => {
return context.dataSources.users.findById(args.id);
},
users: async (parent, args, context: Context) => {
return context.dataSources.users.findAll();
},
},
};
const server = new ApolloServer({
schema: buildSubgraphSchema([{ typeDefs, resolvers }]),
});
```
## Entity Keys and References
```graphql
# products-subgraph/schema.graphql
extend schema
@link(url: "https://specs.apollo.dev/federation/v2.5", import: [
"@key",
"@shareable",
"@interfaceObject"
])
# Single key field
type Product @key(fields: "id") {
id: ID!
name: String!
price: Float!
sku: String! @shareable
}
# Composite key
type Variant @key(fields: "productId sku") {
productId: ID!
sku: String!
size: String!
color: String!
}
# Multiple keys (different ways to identify)
type Review @key(fields: "id") @key(fields: "productId authorId") {
id: ID!
productId: ID!
authorId: ID!
rating: Int!
content: String!
}
```
## Extending Types Across Subgraphs
```graphql
# users-subgraph: owns User
type User @key(fields: "id") {
id: ID!
email: String!
username: String!
}
# posts-subgraph: extends User with posts
extend type User @key(fields: "id") {
id: ID! @external
posts: [Post!]!
}
type Post @key(fields: "id") {
id: ID!
title: String!
content: String!
authorId: ID!
author: User!
}
```
```typescript
// posts-subgraph/resolvers.ts
const resolvers = {
User: {
// Reference resolver: fetch User stub by id
__resolveReference: async (
reference: { id: string },
context: Context
) => {
return { id: reference.id };
},
// Field resolver: resolve posts for User
posts: async (user: { id: string }, args, context: Context) => {
return context.dataSources.posts.findByAuthor(user.id);
},
},
Post: {
// Resolve author as User entity reference
author: (post: Post) => {
return { __typename: 'User', id: post.authorId };
},
},
};
```
## Federation Directives
```graphql
extend schema
@link(url: "https://specs.apollo.dev/federation/v2.5", import: [
"@key",
"@requires",
"@provides",
"@external",
"@shareable",
"@override",
"@inaccessible",
"@tag"
])
# @key: Define entity with primary key
type Product @key(fields: "id") {
id: ID!
name: String!
}
# @external: Field defined in another subgraph
extend type User @key(fields: "id") {
id: ID! @external
email: String! @external
isVerified: Boolean! @external
}
# @requires: Field needs external data
extend type User @key(fields: "id") {
id: ID! @external
email: String! @external
isVerified: Boolean! @external
# Can only compute if we have email and isVerified
canPost: Boolean! @requires(fields: "email isVerified")
}
# @provides: Optimization hint
type Post @key(fields: "id") {
id: ID!
author: User! @provides(fields: "username")
}
# @shareable: Field can be resolved by multiple subgraphs
type Product @key(fields: "id") {
id: ID!
sku: String! @shareable
name: String!
}
# @override: Migration between subgraphs
type Product @key(fields: "id") {
id: ID!
# Override from legacy-subgraph
price: Float! @override(from: "legacy-subgraph")
}
# @inaccessible: Hide from supergraph
type User @key(fields: "id") {
id: ID!
email: String!
internalId: String! @inaccessible
}
# @tag: Organize schema
type Query {
products: [Product!]! @tag(name: "public")
adminUsers: [User!]! @tag(name: "admin")
}
```
## Gateway Configuration
```typescript
// gateway/server.ts
import { ApolloGateway, IntrospectAndCompose } from '@apollo/gateway';
import { ApolloServer } from '@apollo/server';
const gateway = new ApolloGateway({
supergraphSdl: new IntrospectAndCompose({
subgraphs: [
{ name: 'users', url: 'http://localhost:4001/graphql' },
{ name: 'posts', url: 'http://localhost:4002/graphql' },
{ name: 'products', url: 'http://localhost:4003/graphql' },
],
// Poll for schema updates
pollIntervalInMs: 10000,
}),
// Error handling
serviceHealthCheck: true,
// Query planning debug
debug: process.env.NODE_ENV === 'development',
});
const server = new ApolloServer({
gateway,
// Context propagation to subgraphs
context: async ({ req }) => {
const token = req.headers.authorization || '';
return { token };
},
});
await server.listen(4000);
console.log('Gateway ready at http://localhost:4000');
```
## Managed Federation (Apollo Studio)
```typescript
// gateway/server.ts with managed federation
import { ApolloGateway } from '@apollo/gateway';
import { ApolloServer } from '@apollo/server';
const gateway = new ApolloGateway({
// No subgraph URLs needed - fetched from Apollo Studio
// Schema composition happens in Apollo Studio
async supergraphSdl({ update }) {
// Fetch from Apollo Uplink
const supergraphSdl = await fetchSupergraphSdl();
return {
supergraphSdl,
cleanup: async () => {},
};
},
});
// Subgraph reporting to Apollo Studio
import { ApolloServerPluginInlineTrace } from '@apollo/server/plugin/inlineTrace';
const subgraphServer = new ApolloServer({
schema: buildSubgraphSchema([{ typeDefs, resolvers }]),
plugins: [
ApolloServerPluginInlineTrace(),
],
});
```
## Value Types vs Entities
```graphql
# Value type: no @key, resolved entirely by one subgraph
type Address {
street: String!
city: String!
country: String!
postalCode: String!
}
# Entity: has @key, can be extended by other subgraphs
type User @key(fields: "id") {
id: ID!
email: String!
# Value type embedded in entity
address: Address
}
# Another subgraph can extend User but not Address
extend type User @key(fields: "id") {
id: ID! @external
orders: [Order!]!
}
```
## Interface Objects
```graphql
# accounts-subgraph
type User implements Account @key(fields: "id") {
id: ID!
email: String!
role: String!
}
type AdminUser implements Account @key(fields: "id") {
id: ID!
email: String!
role: String!
permissions: [String!]!
}
interface Account {
id: ID!
email: String!
role: String!
}
# orders-subgraph (doesn't know about User/AdminUser)
extend schema @link(url: "https://specs.apollo.dev/federation/v2.5", import: ["@key", "@interfaceObject"])
type Order @key(fields: "id") {
id: ID!
account: Account!
}
# Use @interfaceObject to reference Account without knowing implementations
type Account @key(fields: "id") @interfaceObject {
id: ID!
}
```
## Query Planning Optimization
```graphql
# Inefficient: requires multiple roundtrips
type Query {
user(id: ID!): User
}
type User @key(fields: "id") {
id: ID!
posts: [Post!]!
}
extend type Post @key(fields: "id") {
id: ID! @external
author: User!
}
# Better: provide data to avoid extra fetch
type Post @key(fields: "id") {
id: ID!
authorId: ID!
# Optimization: provide username directly
author: User! @provides(fields: "username")
}
# Gateway can fulfill some User fields from Post subgraph
# without fetching from User subgraph
```
## Error Handling in Federation
```typescript
const resolvers = {
User: {
__resolveReference: async (
reference: { id: string },
context: Context
) => {
try {
const user = await context.dataSources.users.findById(reference.id);
if (!user) {
// Return null for missing entity (soft error)
return null;
}
return user;
} catch (error) {
// Hard error propagates to client
throw new GraphQLError('Failed to resolve user', {
extensions: {
code: 'USER_RESOLUTION_FAILED',
userId: reference.id,
},
});
}
},
},
};
```
## Federation Best Practices
1. **Entity Design**: Use @key for types that need to be extended
2. **Subgraph Boundaries**: Align with team/service boundaries
3. **Shared Types**: Use @shareable for truly shared fields
4. **Migration**: Use @override for gradual subgraph migration
5. **Performance**: Use @provides to optimize query planning
6. **Value Types**: Use plain types for embedded data
7. **Composition**: Test schema composition in CI/CD
8. **Versioning**: Use managed federation for safe deployments
9. **Monitoring**: Track query planning and resolver performance
10. **Documentation**: Document entity ownership and extension patterns

1141
skills/graphql-architect/references/migration-from-rest.md
View File

File diff suppressed because it is too large Load Diff

425
skills/graphql-architect/references/resolvers.md
View File

@ -1,425 +0,0 @@
# GraphQL Resolvers
## Basic Resolver Pattern
```typescript
import { GraphQLResolveInfo } from 'graphql';
// Resolver signature
type Resolver<TSource, TArgs, TContext, TReturn> = (
parent: TSource,
args: TArgs,
context: TContext,
info: GraphQLResolveInfo
) => Promise<TReturn> | TReturn;
// User resolvers
const resolvers = {
Query: {
user: async (
parent,
args: { id: string },
context: Context
): Promise<User | null> => {
return context.dataSources.users.findById(args.id);
},
users: async (
parent,
args: { first?: number; after?: string },
context: Context
): Promise<User[]> => {
return context.dataSources.users.findAll(args);
},
},
Mutation: {
createUser: async (
parent,
args: { input: CreateUserInput },
context: Context
): Promise<User> => {
if (!context.user) {
throw new Error('Unauthorized');
}
return context.dataSources.users.create(args.input);
},
},
};
```
## Context Setup
```typescript
import { Request } from 'express';
import { User } from './models';
import { DataSources } from './datasources';
export interface Context {
user: User | null;
dataSources: DataSources;
loaders: Loaders;
req: Request;
authToken: string | null;
}
// Apollo Server context
const server = new ApolloServer({
typeDefs,
resolvers,
context: async ({ req }): Promise<Context> => {
// Extract auth token
const authToken = req.headers.authorization?.replace('Bearer ', '') || null;
// Verify user
let user: User | null = null;
if (authToken) {
user = await verifyToken(authToken);
}
// Create data sources
const dataSources = new DataSources({
db: prisma,
redis: redisClient,
});
// Create DataLoaders
const loaders = createLoaders(dataSources);
return {
user,
dataSources,
loaders,
req,
authToken,
};
},
});
```
## DataLoader for N+1 Prevention
```typescript
import DataLoader from 'dataloader';
// Create loaders
export function createLoaders(dataSources: DataSources): Loaders {
return {
userLoader: new DataLoader<string, User>(
async (ids: readonly string[]) => {
const users = await dataSources.users.findByIds([...ids]);
// Return in same order as input ids
return ids.map(id => users.find(u => u.id === id) || null);
},
{
cache: true,
batchScheduleFn: (callback) => setTimeout(callback, 10),
}
),
postsByAuthorLoader: new DataLoader<string, Post[]>(
async (authorIds: readonly string[]) => {
const posts = await dataSources.posts.findByAuthorIds([...authorIds]);
// Group by author
return authorIds.map(authorId =>
posts.filter(p => p.authorId === authorId)
);
}
),
};
}
// Field resolver using DataLoader
const resolvers = {
Post: {
author: async (
post: Post,
args,
context: Context
): Promise<User> => {
// Batches multiple requests into single DB query
return context.loaders.userLoader.load(post.authorId);
},
},
User: {
posts: async (
user: User,
args,
context: Context
): Promise<Post[]> => {
return context.loaders.postsByAuthorLoader.load(user.id);
},
},
};
```
## Field Resolvers
```typescript
const resolvers = {
User: {
// Simple field resolver
fullName: (user: User): string => {
return `${user.firstName} ${user.lastName}`;
},
// Async field resolver with DB query
postCount: async (
user: User,
args,
context: Context
): Promise<number> => {
return context.dataSources.posts.countByAuthor(user.id);
},
// Field resolver with arguments
posts: async (
user: User,
args: { first?: number; status?: PostStatus },
context: Context
): Promise<Post[]> => {
return context.dataSources.posts.findByAuthor(user.id, {
limit: args.first,
status: args.status,
});
},
// Nullable field with conditional logic
profile: async (
user: User,
args,
context: Context
): Promise<Profile | null> => {
if (!user.hasProfile) return null;
return context.loaders.profileLoader.load(user.id);
},
},
};
```
## Interface Resolvers
```typescript
const resolvers = {
// Interface type resolver
Searchable: {
__resolveType(obj: Article | Video | Podcast): string {
if ('content' in obj) return 'Article';
if ('duration' in obj) return 'Video';
if ('audioUrl' in obj) return 'Podcast';
throw new Error('Unknown Searchable type');
},
},
// Common interface fields (shared resolvers)
Article: {
id: (article: Article) => article.id,
title: (article: Article) => article.title,
description: (article: Article) => article.description,
},
Video: {
id: (video: Video) => video.id,
title: (video: Video) => video.title,
description: (video: Video) => video.description,
},
};
```
## Union Resolvers
```typescript
const resolvers = {
// Union type resolver
SearchResult: {
__resolveType(
obj: Article | Video | Podcast,
context: Context,
info: GraphQLResolveInfo
): string {
if ('content' in obj) return 'Article';
if ('duration' in obj && 'url' in obj) return 'Video';
if ('audioUrl' in obj) return 'Podcast';
throw new Error('Unknown SearchResult type');
},
},
Query: {
searchContent: async (
parent,
args: { query: string },
context: Context
): Promise<(Article | Video | Podcast)[]> => {
// Return mixed array of different types
const [articles, videos, podcasts] = await Promise.all([
context.dataSources.articles.search(args.query),
context.dataSources.videos.search(args.query),
context.dataSources.podcasts.search(args.query),
]);
return [...articles, ...videos, ...podcasts];
},
},
};
```
## Error Handling
```typescript
import { GraphQLError } from 'graphql';
import { ApolloServerErrorCode } from '@apollo/server/errors';
const resolvers = {
Query: {
user: async (
parent,
args: { id: string },
context: Context
): Promise<User> => {
const user = await context.dataSources.users.findById(args.id);
if (!user) {
throw new GraphQLError('User not found', {
extensions: {
code: 'USER_NOT_FOUND',
http: { status: 404 },
userId: args.id,
},
});
}
return user;
},
},
Mutation: {
updateUser: async (
parent,
args: { id: string; input: UpdateUserInput },
context: Context
): Promise<User> => {
// Check authentication
if (!context.user) {
throw new GraphQLError('Unauthorized', {
extensions: {
code: ApolloServerErrorCode.UNAUTHENTICATED,
http: { status: 401 },
},
});
}
// Check authorization
if (context.user.id !== args.id && !context.user.isAdmin) {
throw new GraphQLError('Forbidden', {
extensions: {
code: ApolloServerErrorCode.FORBIDDEN,
http: { status: 403 },
},
});
}
try {
return await context.dataSources.users.update(args.id, args.input);
} catch (error) {
throw new GraphQLError('Failed to update user', {
extensions: {
code: 'UPDATE_FAILED',
originalError: error,
},
});
}
},
},
};
```
## Pagination Resolvers
```typescript
import { encodeCursor, decodeCursor } from './utils/cursor';
const resolvers = {
Query: {
posts: async (
parent,
args: { first?: number; after?: string },
context: Context
): Promise<PostConnection> => {
const limit = Math.min(args.first || 10, 100);
const cursor = args.after ? decodeCursor(args.after) : null;
// Fetch one extra to determine hasNextPage
const posts = await context.dataSources.posts.findAll({
limit: limit + 1,
cursor,
});
const hasNextPage = posts.length > limit;
const edges = posts.slice(0, limit).map(post => ({
node: post,
cursor: encodeCursor(post.id),
}));
return {
edges,
pageInfo: {
hasNextPage,
hasPreviousPage: !!cursor,
startCursor: edges[0]?.cursor || null,
endCursor: edges[edges.length - 1]?.cursor || null,
},
totalCount: await context.dataSources.posts.count(),
};
},
},
};
```
## Batching Patterns
```typescript
// Batch multiple queries
class UserDataSource {
private db: PrismaClient;
async findByIds(ids: string[]): Promise<User[]> {
// Single query instead of N queries
return this.db.user.findMany({
where: { id: { in: ids } },
});
}
async findByEmails(emails: string[]): Promise<User[]> {
return this.db.user.findMany({
where: { email: { in: emails } },
});
}
}
// DataLoader with caching
const userLoader = new DataLoader<string, User>(
async (ids) => {
console.log('Batching user queries:', ids.length);
const users = await dataSources.users.findByIds([...ids]);
return ids.map(id => users.find(u => u.id === id) || null);
},
{
cache: true,
maxBatchSize: 100,
batchScheduleFn: (callback) => setTimeout(callback, 10),
}
);
```
## Resolver Best Practices
1. **Use DataLoader**: Always batch and cache database queries
2. **Avoid N+1**: Use DataLoader for all foreign key relationships
3. **Type Safety**: Use TypeScript for resolver type safety
4. **Error Handling**: Throw GraphQLError with proper codes and extensions
5. **Authorization**: Check permissions in resolvers, not data sources
6. **Pagination**: Implement cursor-based pagination for lists
7. **Context**: Keep context creation lightweight
8. **Caching**: Use DataLoader caching per request
9. **Batching**: Batch queries with DataLoader or in data source
10. **Testing**: Unit test resolvers with mocked context

393
skills/graphql-architect/references/schema-design.md
View File

@ -1,393 +0,0 @@
# GraphQL Schema Design
## Object Types
```graphql
"""
User account with authentication and profile information.
All users must have a unique email address.
"""
type User {
"Unique user identifier"
id: ID!
"User's email address (unique)"
email: String!
"Display name (optional)"
username: String
"Account creation timestamp"
createdAt: DateTime!
"User's posts (paginated)"
posts(first: Int = 10, after: String): PostConnection!
"User's profile (nullable if not completed)"
profile: Profile
}
type Profile {
id: ID!
bio: String
avatarUrl: URL
website: URL
location: String
}
type Post {
id: ID!
title: String!
content: String!
author: User!
publishedAt: DateTime
status: PostStatus!
tags: [Tag!]!
comments(first: Int, after: String): CommentConnection!
}
```
## Interfaces
```graphql
"""
Common interface for all content that can be timestamped
"""
interface Timestamped {
id: ID!
createdAt: DateTime!
updatedAt: DateTime!
}
"""
Interface for searchable content
"""
interface Searchable {
id: ID!
title: String!
description: String
}
type Article implements Timestamped & Searchable {
id: ID!
title: String!
description: String
content: String!
createdAt: DateTime!
updatedAt: DateTime!
author: User!
}
type Video implements Timestamped & Searchable {
id: ID!
title: String!
description: String
url: URL!
duration: Int!
createdAt: DateTime!
updatedAt: DateTime!
uploader: User!
}
# Query returning interface
type Query {
search(query: String!): [Searchable!]!
}
```
## Union Types
```graphql
"""
Result of a content search - can be Article, Video, or Podcast
"""
union SearchResult = Article | Video | Podcast
"""
Notification types that users can receive
"""
union Notification = CommentNotification | LikeNotification | FollowNotification
type CommentNotification {
id: ID!
comment: Comment!
post: Post!
createdAt: DateTime!
}
type LikeNotification {
id: ID!
liker: User!
post: Post!
createdAt: DateTime!
}
type Query {
searchContent(query: String!): [SearchResult!]!
notifications(first: Int): [Notification!]!
}
```
## Enums
```graphql
"""
Post publication status
"""
enum PostStatus {
DRAFT
PUBLISHED
ARCHIVED
DELETED
}
"""
User role for authorization
"""
enum UserRole {
ADMIN
MODERATOR
USER
GUEST
}
"""
Sort direction for queries
"""
enum SortOrder {
ASC
DESC
}
type Query {
posts(
status: PostStatus
orderBy: SortOrder = DESC
): [Post!]!
}
```
## Input Types
```graphql
"""
Input for creating a new user
"""
input CreateUserInput {
email: String!
password: String!
username: String
profile: ProfileInput
}
input ProfileInput {
bio: String
avatarUrl: URL
website: URL
location: String
}
"""
Input for updating a post
"""
input UpdatePostInput {
title: String
content: String
status: PostStatus
tags: [ID!]
}
"""
Pagination and filtering input
"""
input PostFilterInput {
status: PostStatus
authorId: ID
tags: [String!]
search: String
createdAfter: DateTime
createdBefore: DateTime
}
type Mutation {
createUser(input: CreateUserInput!): User!
updatePost(id: ID!, input: UpdatePostInput!): Post!
}
type Query {
posts(filter: PostFilterInput, first: Int, after: String): PostConnection!
}
```
## Custom Scalars
```graphql
"""
ISO 8601 date-time string
"""
scalar DateTime
"""
Valid URL string
"""
scalar URL
"""
Valid email address
"""
scalar Email
"""
JSON object
"""
scalar JSON
"""
Positive integer
"""
scalar PositiveInt
type User {
id: ID!
email: Email!
createdAt: DateTime!
website: URL
metadata: JSON
age: PositiveInt
}
```
## Pagination Patterns
```graphql
"""
Cursor-based pagination (Relay specification)
"""
type PostConnection {
edges: [PostEdge!]!
pageInfo: PageInfo!
totalCount: Int!
}
type PostEdge {
node: Post!
cursor: String!
}
type PageInfo {
hasNextPage: Boolean!
hasPreviousPage: Boolean!
startCursor: String
endCursor: String
}
type Query {
posts(
first: Int
after: String
last: Int
before: String
): PostConnection!
}
```
## Nullable vs Non-Nullable Best Practices
```graphql
type User {
# Non-nullable: guaranteed to exist
id: ID!
email: String!
createdAt: DateTime!
# Nullable: optional or may not exist yet
username: String
bio: String
avatarUrl: URL
# Non-null list of nullable items
# List always exists but can be empty, items can be null
tags: [String]!
# Non-null list of non-null items
# List always exists, all items guaranteed non-null
roles: [UserRole!]!
# Nullable list of non-null items
# List may be null, but if exists, all items non-null
posts: [Post!]
}
type Query {
# Non-null: query always returns result (empty list if none)
users: [User!]!
# Nullable: may return null if not found
user(id: ID!): User
# Non-null: guaranteed to return result or error
currentUser: User!
}
```
## Field Deprecation
```graphql
type User {
id: ID!
email: String!
# Deprecated field with migration path
name: String @deprecated(reason: "Use 'username' instead")
username: String
# Deprecated with specific date
legacyId: String @deprecated(
reason: "Migrating to UUID. Will be removed 2025-06-01"
)
}
```
## Schema Documentation
```graphql
"""
User represents an authenticated account in the system.
Users can create posts, comments, and interact with content.
Example query:
```
query GetUser {
user(id: "123") {
email
username
posts(first: 10) {
edges {
node {
title
}
}
}
}
}
```
"""
type User {
"Unique identifier for the user"
id: ID!
"Email address (must be unique across all users)"
email: String!
"Optional display name (defaults to email if not set)"
username: String
}
```
## Design Principles
1. **Nullable Fields**: Make fields nullable by default unless guaranteed to exist
2. **List Fields**: Use `[Type!]!` for lists that always exist with non-null items
3. **Documentation**: Document all types and fields with descriptions
4. **Naming**: Use camelCase for fields, PascalCase for types
5. **Interfaces**: Use interfaces for shared fields across types
6. **Unions**: Use unions for polymorphic return types
7. **Input Types**: Create separate input types for mutations
8. **Scalars**: Use custom scalars for domain-specific types
9. **Deprecation**: Mark deprecated fields, provide migration path
10. **Examples**: Include example queries in documentation

569
skills/graphql-architect/references/security.md
View File

@ -1,569 +0,0 @@
# GraphQL Security
## Query Depth Limiting
```typescript
import depthLimit from 'graphql-depth-limit';
import { ApolloServer } from '@apollo/server';
const server = new ApolloServer({
typeDefs,
resolvers,
validationRules: [
// Limit query depth to 7 levels
depthLimit(7, {
ignore: [
'_service',
'_entities',
'pageInfo',
'edges',
'node',
],
}),
],
});
// Example: This query would be rejected (depth > 7)
// query TooDeep {
// user {
// posts {
// author {
// posts {
// author {
// posts {
// author {
// posts { # Depth 7
// author { # Depth 8 - REJECTED
// name
// }
// }
// }
// }
// }
// }
// }
// }
// }
// }
```
## Query Complexity Analysis
```typescript
import { createComplexityRule } from 'graphql-validation-complexity';
import { GraphQLError } from 'graphql';
// Define field complexities
const complexityRule = createComplexityRule({
maximumComplexity: 1000,
variables: {},
onCost: (cost) => {
console.log('Query cost:', cost);
},
createError(cost, documentNode) {
return new GraphQLError(
`Query too complex: ${cost}. Maximum allowed: 1000`,
{
extensions: {
code: 'COMPLEXITY_LIMIT_EXCEEDED',
cost,
limit: 1000,
},
}
);
},
estimators: [
// Simple field: cost 1
{
estimateComplexity: ({ type }) => {
if (type.toString() === 'String' || type.toString() === 'Int') {
return 1;
}
return 0;
},
},
// List field: cost based on `first` argument
{
estimateComplexity: ({ args, childComplexity }) => {
const first = args.first || 10;
return first * childComplexity;
},
},
],
});
const server = new ApolloServer({
typeDefs,
resolvers,
validationRules: [complexityRule],
});
```
## Custom Complexity Directives
```graphql
# Schema definition
directive @cost(
complexity: Int!
multipliers: [String!]
) on FIELD_DEFINITION
type Query {
# Simple query: cost 1
user(id: ID!): User
# List query: cost multiplied by `first` argument
users(first: Int = 10): [User!]! @cost(complexity: 1, multipliers: ["first"])
# Expensive query: cost 50
analytics: Analytics! @cost(complexity: 50)
}
type User {
id: ID!
name: String! @cost(complexity: 1)
# Related list: cost multiplied by `first`
posts(first: Int = 10): [Post!]! @cost(complexity: 2, multipliers: ["first"])
# Expensive computation
recommendations: [User!]! @cost(complexity: 20)
}
```
```typescript
// Complexity calculator implementation
import { DirectiveNode } from 'graphql';
function calculateComplexity(
field: any,
args: Record<string, any>,
childComplexity: number
): number {
const costDirective = field.astNode?.directives?.find(
(d: DirectiveNode) => d.name.value === 'cost'
);
if (!costDirective) {
return 1 + childComplexity;
}
const complexity =
costDirective.arguments?.find((a) => a.name.value === 'complexity')
?.value.value || 1;
const multipliers =
costDirective.arguments?.find((a) => a.name.value === 'multipliers')
?.value.values || [];
let cost = complexity;
for (const multiplier of multipliers) {
const argValue = args[multiplier.value] || 1;
cost *= argValue;
}
return cost + childComplexity;
}
```
## Rate Limiting
```typescript
import rateLimit from 'express-rate-limit';
import RedisStore from 'rate-limit-redis';
import Redis from 'ioredis';
// IP-based rate limiting
const limiter = rateLimit({
store: new RedisStore({
client: new Redis(),
}),
windowMs: 15 * 60 * 1000, // 15 minutes
max: 100, // 100 requests per window
message: 'Too many requests from this IP',
standardHeaders: true,
legacyHeaders: false,
});
app.use('/graphql', limiter);
// User-based rate limiting (more sophisticated)
import { RateLimiterRedis } from 'rate-limiter-flexible';
const rateLimiter = new RateLimiterRedis({
storeClient: new Redis(),
points: 1000, // Number of points
duration: 60, // Per 60 seconds
blockDuration: 60 * 5, // Block for 5 minutes if exceeded
});
// In context creation
const server = new ApolloServer({
typeDefs,
resolvers,
context: async ({ req }) => {
const userId = getUserId(req);
try {
await rateLimiter.consume(userId, 1);
} catch (error) {
throw new GraphQLError('Rate limit exceeded', {
extensions: {
code: 'RATE_LIMIT_EXCEEDED',
retryAfter: error.msBeforeNext / 1000,
},
});
}
return { userId };
},
});
```
## Authentication
```typescript
import jwt from 'jsonwebtoken';
import { GraphQLError } from 'graphql';
// JWT verification
function verifyToken(token: string): User | null {
try {
const decoded = jwt.verify(token, process.env.JWT_SECRET!);
return decoded as User;
} catch (error) {
return null;
}
}
// Context with authentication
const server = new ApolloServer({
typeDefs,
resolvers,
context: async ({ req }): Promise<Context> => {
const authHeader = req.headers.authorization || '';
const token = authHeader.replace('Bearer ', '');
let user: User | null = null;
if (token) {
user = verifyToken(token);
}
return {
user,
dataSources: createDataSources(),
};
},
});
// Protected resolvers
const resolvers = {
Query: {
me: (parent, args, context: Context) => {
if (!context.user) {
throw new GraphQLError('Unauthorized', {
extensions: { code: 'UNAUTHENTICATED' },
});
}
return context.user;
},
},
Mutation: {
createPost: (parent, args, context: Context) => {
if (!context.user) {
throw new GraphQLError('Unauthorized', {
extensions: { code: 'UNAUTHENTICATED' },
});
}
return context.dataSources.posts.create({
...args.input,
authorId: context.user.id,
});
},
},
};
```
## Authorization Patterns
```typescript
// Directive-based authorization
import { mapSchema, getDirective, MapperKind } from '@graphql-tools/utils';
import { defaultFieldResolver } from 'graphql';
function authDirective(directiveName: string) {
return (schema: GraphQLSchema) =>
mapSchema(schema, {
[MapperKind.OBJECT_FIELD]: (fieldConfig) => {
const authDirective = getDirective(
schema,
fieldConfig,
directiveName
)?.[0];
if (authDirective) {
const { requires } = authDirective;
const { resolve = defaultFieldResolver } = fieldConfig;
fieldConfig.resolve = async (source, args, context, info) => {
// Check if user has required role
if (!context.user) {
throw new GraphQLError('Unauthorized', {
extensions: { code: 'UNAUTHENTICATED' },
});
}
if (requires && !context.user.roles.includes(requires)) {
throw new GraphQLError('Forbidden', {
extensions: {
code: 'FORBIDDEN',
requiredRole: requires,
},
});
}
return resolve(source, args, context, info);
};
}
return fieldConfig;
},
});
}
// Schema with directives
const typeDefs = gql`
directive @auth(requires: Role) on FIELD_DEFINITION
enum Role {
ADMIN
USER
GUEST
}
type Query {
publicData: String!
userData: String! @auth(requires: USER)
adminData: String! @auth(requires: ADMIN)
}
`;
const schema = authDirective('auth')(makeExecutableSchema({ typeDefs, resolvers }));
```
## Field-Level Authorization
```typescript
// Row-level security
const resolvers = {
Query: {
posts: async (parent, args, context: Context) => {
// Filter based on user permissions
const posts = await context.dataSources.posts.findAll();
return posts.filter((post) => {
// Public posts visible to all
if (post.isPublic) return true;
// Private posts only visible to author
if (context.user?.id === post.authorId) return true;
// Check if user is admin
if (context.user?.roles.includes('ADMIN')) return true;
return false;
});
},
},
Post: {
// Hide email unless viewer is author or admin
authorEmail: (post: Post, args, context: Context) => {
if (!context.user) return null;
if (
context.user.id === post.authorId ||
context.user.roles.includes('ADMIN')
) {
return post.authorEmail;
}
return null;
},
},
};
```
## Query Allowlisting
```typescript
// Persisted queries (automatic allowlisting)
import { createPersistedQueryLink } from '@apollo/client/link/persisted-queries';
import { createHash } from 'crypto';
// Client side
const link = createPersistedQueryLink({
sha256: (query) => createHash('sha256').update(query).digest('hex'),
useGETForHashedQueries: true,
});
// Server side
import { ApolloServerPluginInlineTrace } from '@apollo/server/plugin/inlineTrace';
const server = new ApolloServer({
typeDefs,
resolvers,
persistedQueries: {
cache: new Map(), // or Redis
},
// Only allow persisted queries in production
allowBatchedHttpRequests: false,
introspection: process.env.NODE_ENV !== 'production',
});
// Manual allowlist
const allowedOperations = new Set([
'GetUser',
'GetPosts',
'CreatePost',
'UpdatePost',
]);
const server = new ApolloServer({
typeDefs,
resolvers,
plugins: [
{
async requestDidStart() {
return {
async didResolveOperation(requestContext) {
const operationName = requestContext.operationName;
if (!operationName || !allowedOperations.has(operationName)) {
throw new GraphQLError('Operation not allowed', {
extensions: { code: 'OPERATION_NOT_ALLOWED' },
});
}
},
};
},
},
],
});
```
## Input Validation
```typescript
import { z } from 'zod';
// Zod schema for input validation
const CreatePostSchema = z.object({
title: z.string().min(3).max(200),
content: z.string().min(10).max(10000),
tags: z.array(z.string()).max(5),
isPublic: z.boolean(),
});
const resolvers = {
Mutation: {
createPost: async (
parent,
args: { input: any },
context: Context
) => {
// Validate input
const validationResult = CreatePostSchema.safeParse(args.input);
if (!validationResult.success) {
throw new GraphQLError('Invalid input', {
extensions: {
code: 'BAD_USER_INPUT',
validationErrors: validationResult.error.errors,
},
});
}
const input = validationResult.data;
return context.dataSources.posts.create(input);
},
},
};
```
## Introspection Control
```typescript
// Disable introspection in production
import { ApolloServer } from '@apollo/server';
import { ApolloServerPluginLandingPageDisabled } from '@apollo/server/plugin/disabled';
const server = new ApolloServer({
typeDefs,
resolvers,
introspection: process.env.NODE_ENV !== 'production',
plugins:
process.env.NODE_ENV === 'production'
? [ApolloServerPluginLandingPageDisabled()]
: [],
});
// Conditional introspection (admin only)
const server = new ApolloServer({
typeDefs,
resolvers,
introspection: false, // Disable by default
plugins: [
{
async requestDidStart({ request, contextValue }) {
// Allow introspection for admins
if (
request.operationName === 'IntrospectionQuery' &&
!contextValue.user?.isAdmin
) {
throw new GraphQLError('Introspection disabled', {
extensions: { code: 'FORBIDDEN' },
});
}
},
},
],
});
```
## CSRF Protection
```typescript
import csrf from 'csurf';
// CSRF protection for mutations
const csrfProtection = csrf({ cookie: true });
app.post('/graphql', csrfProtection, expressMiddleware(server));
// Client must send CSRF token
// fetch('/graphql', {
// method: 'POST',
// headers: {
// 'CSRF-Token': csrfToken,
// },
// body: JSON.stringify({ query }),
// });
```
## Security Best Practices
1. **Depth Limiting**: Prevent deeply nested queries
2. **Complexity Analysis**: Calculate and limit query cost
3. **Rate Limiting**: Limit requests per user/IP
4. **Authentication**: Verify user identity in context
5. **Authorization**: Check permissions in resolvers
6. **Input Validation**: Validate all mutation inputs
7. **Query Allowlisting**: Use persisted queries in production
8. **Introspection Control**: Disable in production
9. **Error Sanitization**: Don't expose sensitive data in errors
10. **CORS Configuration**: Restrict allowed origins
11. **HTTPS Only**: Always use HTTPS in production
12. **Audit Logging**: Log sensitive operations

510
skills/graphql-architect/references/subscriptions.md
View File

@ -1,510 +0,0 @@
# GraphQL Subscriptions
## Basic Subscription Setup
```typescript
// schema.graphql
type Subscription {
postCreated: Post!
postUpdated(id: ID!): Post!
commentAdded(postId: ID!): Comment!
userOnline: User!
}
type Post {
id: ID!
title: String!
content: String!
author: User!
}
type Comment {
id: ID!
content: String!
author: User!
post: Post!
}
// server.ts
import { createServer } from 'http';
import { ApolloServer } from '@apollo/server';
import { expressMiddleware } from '@apollo/server/express4';
import { ApolloServerPluginDrainHttpServer } from '@apollo/server/plugin/drainHttpServer';
import { makeExecutableSchema } from '@graphql-tools/schema';
import { WebSocketServer } from 'ws';
import { useServer } from 'graphql-ws/lib/use/ws';
import express from 'express';
const schema = makeExecutableSchema({ typeDefs, resolvers });
const app = express();
const httpServer = createServer(app);
// WebSocket server for subscriptions
const wsServer = new WebSocketServer({
server: httpServer,
path: '/graphql',
});
const serverCleanup = useServer(
{
schema,
context: async (ctx, msg, args) => {
// Extract auth from connection params
const token = ctx.connectionParams?.authorization;
const user = token ? await verifyToken(token) : null;
return { user };
},
},
wsServer
);
const server = new ApolloServer({
schema,
plugins: [
ApolloServerPluginDrainHttpServer({ httpServer }),
{
async serverWillStart() {
return {
async drainServer() {
await serverCleanup.dispose();
},
};
},
},
],
});
await server.start();
app.use('/graphql', express.json(), expressMiddleware(server));
httpServer.listen(4000);
```
## PubSub Implementation
```typescript
// pubsub.ts
import { RedisPubSub } from 'graphql-redis-subscriptions';
import Redis from 'ioredis';
// In-memory (development only)
import { PubSub } from 'graphql-subscriptions';
export const pubsub = new PubSub();
// Redis (production)
const options = {
host: process.env.REDIS_HOST || 'localhost',
port: parseInt(process.env.REDIS_PORT || '6379'),
retryStrategy: (times: number) => Math.min(times * 50, 2000),
};
export const pubsub = new RedisPubSub({
publisher: new Redis(options),
subscriber: new Redis(options),
});
// Strongly typed event names
export const EVENTS = {
POST_CREATED: 'POST_CREATED',
POST_UPDATED: 'POST_UPDATED',
COMMENT_ADDED: 'COMMENT_ADDED',
USER_ONLINE: 'USER_ONLINE',
} as const;
```
## Subscription Resolvers
```typescript
import { withFilter } from 'graphql-subscriptions';
import { pubsub, EVENTS } from './pubsub';
const resolvers = {
Subscription: {
// Simple subscription
postCreated: {
subscribe: () => pubsub.asyncIterator([EVENTS.POST_CREATED]),
},
// Filtered subscription
postUpdated: {
subscribe: withFilter(
() => pubsub.asyncIterator([EVENTS.POST_UPDATED]),
(payload, variables) => {
// Only send updates for specific post
return payload.postUpdated.id === variables.id;
}
),
},
// Filtered with authorization
commentAdded: {
subscribe: withFilter(
(parent, args, context) => {
// Check auth before subscribing
if (!context.user) {
throw new Error('Unauthorized');
}
return pubsub.asyncIterator([EVENTS.COMMENT_ADDED]);
},
async (payload, variables, context) => {
// Filter by post and check permissions
if (payload.commentAdded.postId !== variables.postId) {
return false;
}
// Check if user has access to post
const post = await context.dataSources.posts.findById(
variables.postId
);
return post && post.isPublic || post.authorId === context.user.id;
}
),
},
// Complex subscription with multiple filters
userOnline: {
subscribe: withFilter(
() => pubsub.asyncIterator([EVENTS.USER_ONLINE]),
(payload, variables, context) => {
// Only notify friends
return context.user.friends.includes(payload.userOnline.id);
}
),
},
},
Mutation: {
createPost: async (parent, args, context) => {
const post = await context.dataSources.posts.create(args.input);
// Publish event
await pubsub.publish(EVENTS.POST_CREATED, {
postCreated: post,
});
return post;
},
updatePost: async (parent, args: { id: string; input: any }, context) => {
const post = await context.dataSources.posts.update(
args.id,
args.input
);
await pubsub.publish(EVENTS.POST_UPDATED, {
postUpdated: post,
});
return post;
},
addComment: async (parent, args, context) => {
const comment = await context.dataSources.comments.create(args.input);
await pubsub.publish(EVENTS.COMMENT_ADDED, {
commentAdded: comment,
});
return comment;
},
},
};
```
## Advanced Filtering
```typescript
// Type-safe payload
interface PostCreatedPayload {
postCreated: Post;
tags: string[];
isPublic: boolean;
}
const resolvers = {
Subscription: {
postCreated: {
subscribe: withFilter(
() => pubsub.asyncIterator([EVENTS.POST_CREATED]),
async (
payload: PostCreatedPayload,
variables: { tags?: string[]; authorId?: string },
context: Context
) => {
// Filter by tags
if (variables.tags && variables.tags.length > 0) {
const hasMatchingTag = payload.tags.some(tag =>
variables.tags!.includes(tag)
);
if (!hasMatchingTag) return false;
}
// Filter by author
if (variables.authorId) {
if (payload.postCreated.authorId !== variables.authorId) {
return false;
}
}
// Check permissions
if (!payload.isPublic) {
return (
context.user?.id === payload.postCreated.authorId ||
context.user?.isAdmin
);
}
return true;
}
),
},
},
};
```
## Connection Management
```typescript
import { useServer } from 'graphql-ws/lib/use/ws';
const wsServer = useServer(
{
schema,
// Connection lifecycle
onConnect: async (ctx) => {
console.log('Client connected');
const token = ctx.connectionParams?.authorization;
if (!token) {
throw new Error('Missing auth token');
}
const user = await verifyToken(token);
if (!user) {
throw new Error('Invalid token');
}
return { user };
},
onDisconnect: (ctx, code, reason) => {
console.log('Client disconnected', code, reason);
},
// Subscription lifecycle
onSubscribe: async (ctx, msg) => {
console.log('Client subscribed', msg.payload.operationName);
// Rate limiting
const subscriptionCount = getUserSubscriptionCount(ctx.user.id);
if (subscriptionCount >= 10) {
throw new Error('Too many subscriptions');
}
return { ctx, msg };
},
onComplete: (ctx, msg) => {
console.log('Subscription completed', msg.id);
},
// Keep-alive
connectionInitWaitTimeout: 10000,
// Context per subscription
context: async (ctx, msg, args) => {
const user = ctx.extra.user;
return {
user,
dataSources: createDataSources(),
subscriptionId: msg.id,
};
},
},
wsServer
);
```
## Subscription Patterns
```typescript
// Pattern 1: Entity updates
type Subscription {
entityUpdated(id: ID!): Entity!
}
// Pattern 2: Collection updates
type Subscription {
entityAdded: Entity!
entityDeleted: ID!
}
// Pattern 3: Stream of events
type Subscription {
events(types: [EventType!]): Event!
}
// Pattern 4: Live query (with intervals)
type Subscription {
liveQuery(query: String!): [SearchResult!]!
}
// resolvers.ts
const resolvers = {
Subscription: {
// Live query implementation
liveQuery: {
subscribe: async function* (parent, args, context) {
while (true) {
const results = await context.dataSources.search(args.query);
yield { liveQuery: results };
await new Promise(resolve => setTimeout(resolve, 5000));
}
},
},
},
};
```
## Error Handling
```typescript
const resolvers = {
Subscription: {
postCreated: {
subscribe: withFilter(
() => pubsub.asyncIterator([EVENTS.POST_CREATED]),
async (payload, variables, context) => {
try {
// Check permissions
if (!context.user) {
throw new GraphQLError('Unauthorized', {
extensions: { code: 'UNAUTHENTICATED' },
});
}
return true;
} catch (error) {
// Log error but don't propagate to client
console.error('Subscription filter error:', error);
return false;
}
}
),
// Resolve subscription payload
resolve: (payload) => {
try {
return payload.postCreated;
} catch (error) {
throw new GraphQLError('Failed to resolve subscription', {
extensions: { code: 'SUBSCRIPTION_RESOLVE_ERROR' },
});
}
},
},
},
};
```
## Client Usage
```typescript
// Apollo Client setup
import { ApolloClient, InMemoryCache, split, HttpLink } from '@apollo/client';
import { GraphQLWsLink } from '@apollo/client/link/subscriptions';
import { getMainDefinition } from '@apollo/client/utilities';
import { createClient } from 'graphql-ws';
const httpLink = new HttpLink({
uri: 'http://localhost:4000/graphql',
});
const wsLink = new GraphQLWsLink(
createClient({
url: 'ws://localhost:4000/graphql',
connectionParams: {
authorization: `Bearer ${token}`,
},
})
);
const splitLink = split(
({ query }) => {
const definition = getMainDefinition(query);
return (
definition.kind === 'OperationDefinition' &&
definition.operation === 'subscription'
);
},
wsLink,
httpLink
);
const client = new ApolloClient({
link: splitLink,
cache: new InMemoryCache(),
});
// Subscribe to events
const subscription = client
.subscribe({
query: gql`
subscription OnPostCreated {
postCreated {
id
title
author {
username
}
}
}
`,
})
.subscribe({
next: (data) => console.log('New post:', data),
error: (err) => console.error('Subscription error:', err),
complete: () => console.log('Subscription completed'),
});
// Unsubscribe
subscription.unsubscribe();
```
## Scaling Subscriptions
```typescript
// Use Redis for multi-instance deployments
import { RedisPubSub } from 'graphql-redis-subscriptions';
// Horizontal scaling pattern
const pubsub = new RedisPubSub({
publisher: new Redis(redisConfig),
subscriber: new Redis(redisConfig),
// Channel prefix for isolation
publisherPrefix: 'graphql:pub:',
subscriberPrefix: 'graphql:sub:',
});
// Connection limit per instance
const MAX_CONNECTIONS_PER_INSTANCE = 10000;
// Load balancing with sticky sessions
// Ensure same user connects to same server instance
// for connection state management
```
## Subscription Best Practices
1. **Authentication**: Always validate auth in onConnect and filters
2. **Authorization**: Check permissions in withFilter
3. **Rate Limiting**: Limit subscriptions per user
4. **Filtering**: Use withFilter for server-side filtering
5. **Cleanup**: Always clean up subscriptions on disconnect
6. **Scaling**: Use Redis PubSub for multi-instance deployments
7. **Error Handling**: Gracefully handle errors in filters and resolvers
8. **Testing**: Test subscription lifecycle and filtering
9. **Monitoring**: Track active connections and subscription count
10. **Performance**: Avoid N+1 in subscription resolvers

1199
skills/plan-ceo-review/SKILL.md
View File

File diff suppressed because it is too large Load Diff

838
skills/plan-ceo-review/SKILL.md.tmpl
View File

@ -1,838 +0,0 @@
---
name: plan-ceo-review
version: 1.0.0
description: |
CEO/founder-mode plan review. Rethink the problem, find the 10-star product,
challenge premises, expand scope when it creates a better product. Four modes:
SCOPE EXPANSION (dream big), SELECTIVE EXPANSION (hold scope + cherry-pick
expansions), HOLD SCOPE (maximum rigor), SCOPE REDUCTION (strip to essentials).
Use when asked to "think bigger", "expand scope", "strategy review", "rethink this",
or "is this ambitious enough".
Proactively suggest when the user is questioning scope or ambition of a plan,
or when the plan feels like it could be thinking bigger.
benefits-from: [office-hours]
allowed-tools:
- Read
- Grep
- Glob
- Bash
- AskUserQuestion
- WebSearch
---
{{PREAMBLE}}
{{BASE_BRANCH_DETECT}}
# Mega Plan Review Mode
## Philosophy
You are not here to rubber-stamp this plan. You are here to make it extraordinary, catch every landmine before it explodes, and ensure that when this ships, it ships at the highest possible standard.
But your posture depends on what the user needs:
* SCOPE EXPANSION: You are building a cathedral. Envision the platonic ideal. Push scope UP. Ask "what would make this 10x better for 2x the effort?" You have permission to dream — and to recommend enthusiastically. But every expansion is the user's decision. Present each scope-expanding idea as an AskUserQuestion. The user opts in or out.
* SELECTIVE EXPANSION: You are a rigorous reviewer who also has taste. Hold the current scope as your baseline — make it bulletproof. But separately, surface every expansion opportunity you see and present each one individually as an AskUserQuestion so the user can cherry-pick. Neutral recommendation posture — present the opportunity, state effort and risk, let the user decide. Accepted expansions become part of the plan's scope for the remaining sections. Rejected ones go to "NOT in scope."
* HOLD SCOPE: You are a rigorous reviewer. The plan's scope is accepted. Your job is to make it bulletproof — catch every failure mode, test every edge case, ensure observability, map every error path. Do not silently reduce OR expand.
* SCOPE REDUCTION: You are a surgeon. Find the minimum viable version that achieves the core outcome. Cut everything else. Be ruthless.
* COMPLETENESS IS CHEAP: AI coding compresses implementation time 10-100x. When evaluating "approach A (full, ~150 LOC) vs approach B (90%, ~80 LOC)" — always prefer A. The 70-line delta costs seconds with CC. "Ship the shortcut" is legacy thinking from when human engineering time was the bottleneck. Boil the lake.
Critical rule: In ALL modes, the user is 100% in control. Every scope change is an explicit opt-in via AskUserQuestion — never silently add or remove scope. Once the user selects a mode, COMMIT to it. Do not silently drift toward a different mode. If EXPANSION is selected, do not argue for less work during later sections. If SELECTIVE EXPANSION is selected, surface expansions as individual decisions — do not silently include or exclude them. If REDUCTION is selected, do not sneak scope back in. Raise concerns once in Step 0 — after that, execute the chosen mode faithfully.
Do NOT make any code changes. Do NOT start implementation. Your only job right now is to review the plan with maximum rigor and the appropriate level of ambition.
## Prime Directives
1. Zero silent failures. Every failure mode must be visible — to the system, to the team, to the user. If a failure can happen silently, that is a critical defect in the plan.
2. Every error has a name. Don't say "handle errors." Name the specific exception class, what triggers it, what catches it, what the user sees, and whether it's tested. Catch-all error handling (e.g., catch Exception, rescue StandardError, except Exception) is a code smell — call it out.
3. Data flows have shadow paths. Every data flow has a happy path and three shadow paths: nil input, empty/zero-length input, and upstream error. Trace all four for every new flow.
4. Interactions have edge cases. Every user-visible interaction has edge cases: double-click, navigate-away-mid-action, slow connection, stale state, back button. Map them.
5. Observability is scope, not afterthought. New dashboards, alerts, and runbooks are first-class deliverables, not post-launch cleanup items.
6. Diagrams are mandatory. No non-trivial flow goes undiagrammed. ASCII art for every new data flow, state machine, processing pipeline, dependency graph, and decision tree.
7. Everything deferred must be written down. Vague intentions are lies. TODOS.md or it doesn't exist.
8. Optimize for the 6-month future, not just today. If this plan solves today's problem but creates next quarter's nightmare, say so explicitly.
9. You have permission to say "scrap it and do this instead." If there's a fundamentally better approach, table it. I'd rather hear it now.
## Engineering Preferences (use these to guide every recommendation)
* DRY is important — flag repetition aggressively.
* Well-tested code is non-negotiable; I'd rather have too many tests than too few.
* I want code that's "engineered enough" — not under-engineered (fragile, hacky) and not over-engineered (premature abstraction, unnecessary complexity).
* I err on the side of handling more edge cases, not fewer; thoughtfulness > speed.
* Bias toward explicit over clever.
* Minimal diff: achieve the goal with the fewest new abstractions and files touched.
* Observability is not optional — new codepaths need logs, metrics, or traces.
* Security is not optional — new codepaths need threat modeling.
* Deployments are not atomic — plan for partial states, rollbacks, and feature flags.
* ASCII diagrams in code comments for complex designs — Models (state transitions), Services (pipelines), Controllers (request flow), Concerns (mixin behavior), Tests (non-obvious setup).
* Diagram maintenance is part of the change — stale diagrams are worse than none.
## Cognitive Patterns — How Great CEOs Think
These are not checklist items. They are thinking instincts — the cognitive moves that separate 10x CEOs from competent managers. Let them shape your perspective throughout the review. Don't enumerate them; internalize them.
1. **Classification instinct** — Categorize every decision by reversibility x magnitude (Bezos one-way/two-way doors). Most things are two-way doors; move fast.
2. **Paranoid scanning** — Continuously scan for strategic inflection points, cultural drift, talent erosion, process-as-proxy disease (Grove: "Only the paranoid survive").
3. **Inversion reflex** — For every "how do we win?" also ask "what would make us fail?" (Munger).
4. **Focus as subtraction** — Primary value-add is what to *not* do. Jobs went from 350 products to 10. Default: do fewer things, better.
5. **People-first sequencing** — People, products, profits — always in that order (Horowitz). Talent density solves most other problems (Hastings).
6. **Speed calibration** — Fast is default. Only slow down for irreversible + high-magnitude decisions. 70% information is enough to decide (Bezos).
7. **Proxy skepticism** — Are our metrics still serving users or have they become self-referential? (Bezos Day 1).
8. **Narrative coherence** — Hard decisions need clear framing. Make the "why" legible, not everyone happy.
9. **Temporal depth** — Think in 5-10 year arcs. Apply regret minimization for major bets (Bezos at age 80).
10. **Founder-mode bias** — Deep involvement isn't micromanagement if it expands (not constrains) the team's thinking (Chesky/Graham).
11. **Wartime awareness** — Correctly diagnose peacetime vs wartime. Peacetime habits kill wartime companies (Horowitz).
12. **Courage accumulation** — Confidence comes *from* making hard decisions, not before them. "The struggle IS the job."
13. **Willfulness as strategy** — Be intentionally willful. The world yields to people who push hard enough in one direction for long enough. Most people give up too early (Altman).
14. **Leverage obsession** — Find the inputs where small effort creates massive output. Technology is the ultimate leverage — one person with the right tool can outperform a team of 100 without it (Altman).
15. **Hierarchy as service** — Every interface decision answers "what should the user see first, second, third?" Respecting their time, not prettifying pixels.
16. **Edge case paranoia (design)** — What if the name is 47 chars? Zero results? Network fails mid-action? First-time user vs power user? Empty states are features, not afterthoughts.
17. **Subtraction default** — "As little design as possible" (Rams). If a UI element doesn't earn its pixels, cut it. Feature bloat kills products faster than missing features.
18. **Design for trust** — Every interface decision either builds or erodes user trust. Pixel-level intentionality about safety, identity, and belonging.
When you evaluate architecture, think through the inversion reflex. When you challenge scope, apply focus as subtraction. When you assess timeline, use speed calibration. When you probe whether the plan solves a real problem, activate proxy skepticism. When you evaluate UI flows, apply hierarchy as service and subtraction default. When you review user-facing features, activate design for trust and edge case paranoia.
## Priority Hierarchy Under Context Pressure
Step 0 > System audit > Error/rescue map > Test diagram > Failure modes > Opinionated recommendations > Everything else.
Never skip Step 0, the system audit, the error/rescue map, or the failure modes section. These are the highest-leverage outputs.
## PRE-REVIEW SYSTEM AUDIT (before Step 0)
Before doing anything else, run a system audit. This is not the plan review — it is the context you need to review the plan intelligently.
Run the following commands:
```
git log --oneline -30 # Recent history
git diff <base> --stat # What's already changed
git stash list # Any stashed work
grep -r "TODO\|FIXME\|HACK\|XXX" -l --exclude-dir=node_modules --exclude-dir=vendor --exclude-dir=.git . | head -30
git log --since=30.days --name-only --format="" | sort | uniq -c | sort -rn | head -20 # Recently touched files
```
Then read CLAUDE.md, TODOS.md, and any existing architecture docs.
**Design doc check:**
```bash
SLUG=$(~/.claude/skills/gstack/browse/bin/remote-slug 2>/dev/null || basename "$(git rev-parse --show-toplevel 2>/dev/null || pwd)")
BRANCH=$(git rev-parse --abbrev-ref HEAD 2>/dev/null | tr '/' '-' || echo 'no-branch')
DESIGN=$(ls -t ~/.gstack/projects/$SLUG/*-$BRANCH-design-*.md 2>/dev/null | head -1)
[ -z "$DESIGN" ] && DESIGN=$(ls -t ~/.gstack/projects/$SLUG/*-design-*.md 2>/dev/null | head -1)
[ -n "$DESIGN" ] && echo "Design doc found: $DESIGN" || echo "No design doc found"
```
If a design doc exists (from `/office-hours`), read it. Use it as the source of truth for the problem statement, constraints, and chosen approach. If it has a `Supersedes:` field, note that this is a revised design.
**Handoff note check** (reuses $SLUG and $BRANCH from the design doc check above):
```bash
HANDOFF=$(ls -t ~/.gstack/projects/$SLUG/*-$BRANCH-ceo-handoff-*.md 2>/dev/null | head -1)
[ -n "$HANDOFF" ] && echo "HANDOFF_FOUND: $HANDOFF" || echo "NO_HANDOFF"
```
If this block runs in a separate shell from the design doc check, recompute $SLUG and $BRANCH first using the same commands from that block.
If a handoff note is found: read it. This contains system audit findings and discussion
from a prior CEO review session that paused so the user could run `/office-hours`. Use it
as additional context alongside the design doc. The handoff note helps you avoid re-asking
questions the user already answered. Do NOT skip any steps — run the full review, but use
the handoff note to inform your analysis and avoid redundant questions.
Tell the user: "Found a handoff note from your prior CEO review session. I'll use that
context to pick up where we left off."
{{BENEFITS_FROM}}
**Handoff note save (BENEFITS_FROM):** If the user chose A (run /office-hours first),
save a handoff context note before they leave. Reuse $SLUG and $BRANCH from the
design doc check block above (they use the same `remote-slug || basename` fallback
that handles repos without an origin remote). Then run:
```bash
mkdir -p ~/.gstack/projects/$SLUG
USER=$(whoami)
DATETIME=$(date +%Y%m%d-%H%M%S)
```
Write to `~/.gstack/projects/$SLUG/$USER-$BRANCH-ceo-handoff-$DATETIME.md`:
```markdown
# CEO Review Handoff Note
Generated by /plan-ceo-review on {date}
Branch: {branch}
Repo: {owner/repo}
## Why I paused
User chose to run /office-hours first (no design doc found).
## System Audit Summary
{Summarize what the system audit found — recent git history, diff scope,
CLAUDE.md key points, TODOS.md relevant items, known pain points}
## Discussion So Far
{Empty — handoff happened before Step 0. Frontend/UI scope detection has not
run yet — it will be assessed when the review resumes.}
```
Tell the user: "Context saved. Run /office-hours in another window. When you come back
and invoke /plan-ceo-review, I'll pick up the context automatically — including the
design doc /office-hours produces."
**Mid-session detection:** During Step 0A (Premise Challenge), if the user can't
articulate the problem, keeps changing the problem statement, answers with "I'm not
sure," or is clearly exploring rather than reviewing — offer `/office-hours`:
> "It sounds like you're still figuring out what to build — that's totally fine, but
> that's what /office-hours is designed for. Want to pause this review and run
> /office-hours first? It'll help you nail down the problem and approach, then come
> back here for the strategic review."
Options: A) Yes, run /office-hours first. B) No, keep going.
If they keep going, proceed normally — no guilt, no re-asking.
**Handoff note save (mid-session):** If the user chose A (run /office-hours first from
mid-session detection), save a handoff context note with the same format above, but
include any Step 0A progress in the "Discussion So Far" section — premises discussed,
problem framing attempts, user answers so far. Use the same bash block to generate the
file path.
Tell the user: "Context saved with your discussion so far. Run /office-hours, then
come back to /plan-ceo-review."
When reading TODOS.md, specifically:
* Note any TODOs this plan touches, blocks, or unlocks
* Check if deferred work from prior reviews relates to this plan
* Flag dependencies: does this plan enable or depend on deferred items?
* Map known pain points (from TODOS) to this plan's scope
Map:
* What is the current system state?
* What is already in flight (other open PRs, branches, stashed changes)?
* What are the existing known pain points most relevant to this plan?
* Are there any FIXME/TODO comments in files this plan touches?
### Retrospective Check
Check the git log for this branch. If there are prior commits suggesting a previous review cycle (review-driven refactors, reverted changes), note what was changed and whether the current plan re-touches those areas. Be MORE aggressive reviewing areas that were previously problematic. Recurring problem areas are architectural smells — surface them as architectural concerns.
### Frontend/UI Scope Detection
Analyze the plan. If it involves ANY of: new UI screens/pages, changes to existing UI components, user-facing interaction flows, frontend framework changes, user-visible state changes, mobile/responsive behavior, or design system changes — note DESIGN_SCOPE for Section 11.
### Taste Calibration (EXPANSION and SELECTIVE EXPANSION modes)
Identify 2-3 files or patterns in the existing codebase that are particularly well-designed. Note them as style references for the review. Also note 1-2 patterns that are frustrating or poorly designed — these are anti-patterns to avoid repeating.
Report findings before proceeding to Step 0.
### Landscape Check
Read ETHOS.md for the Search Before Building framework (the preamble's Search Before Building section has the path). Before challenging scope, understand the landscape. WebSearch for:
- "[product category] landscape {current year}"
- "[key feature] alternatives"
- "why [incumbent/conventional approach] [succeeds/fails]"
If WebSearch is unavailable, skip this check and note: "Search unavailable — proceeding with in-distribution knowledge only."
Run the three-layer synthesis:
- **[Layer 1]** What's the tried-and-true approach in this space?
- **[Layer 2]** What are the search results saying?
- **[Layer 3]** First-principles reasoning — where might the conventional wisdom be wrong?
Feed into the Premise Challenge (0A) and Dream State Mapping (0C). If you find a eureka moment, surface it during the Expansion opt-in ceremony as a differentiation opportunity. Log it (see preamble).
## Step 0: Nuclear Scope Challenge + Mode Selection
### 0A. Premise Challenge
1. Is this the right problem to solve? Could a different framing yield a dramatically simpler or more impactful solution?
2. What is the actual user/business outcome? Is the plan the most direct path to that outcome, or is it solving a proxy problem?
3. What would happen if we did nothing? Real pain point or hypothetical one?
### 0B. Existing Code Leverage
1. What existing code already partially or fully solves each sub-problem? Map every sub-problem to existing code. Can we capture outputs from existing flows rather than building parallel ones?
2. Is this plan rebuilding anything that already exists? If yes, explain why rebuilding is better than refactoring.
### 0C. Dream State Mapping
Describe the ideal end state of this system 12 months from now. Does this plan move toward that state or away from it?
```
CURRENT STATE THIS PLAN 12-MONTH IDEAL
[describe] ---> [describe delta] ---> [describe target]
```
### 0C-bis. Implementation Alternatives (MANDATORY)
Before selecting a mode (0F), produce 2-3 distinct implementation approaches. This is NOT optional — every plan must consider alternatives.
For each approach:
```
APPROACH A: [Name]
Summary: [1-2 sentences]
Effort: [S/M/L/XL]
Risk: [Low/Med/High]
Pros: [2-3 bullets]
Cons: [2-3 bullets]
Reuses: [existing code/patterns leveraged]
APPROACH B: [Name]
...
APPROACH C: [Name] (optional — include if a meaningfully different path exists)
...
```
**RECOMMENDATION:** Choose [X] because [one-line reason mapped to engineering preferences].
Rules:
- At least 2 approaches required. 3 preferred for non-trivial plans.
- One approach must be the "minimal viable" (fewest files, smallest diff).
- One approach must be the "ideal architecture" (best long-term trajectory).
- If only one approach exists, explain concretely why alternatives were eliminated.
- Do NOT proceed to mode selection (0F) without user approval of the chosen approach.
### 0D. Mode-Specific Analysis
**For SCOPE EXPANSION** — run all three, then the opt-in ceremony:
1. 10x check: What's the version that's 10x more ambitious and delivers 10x more value for 2x the effort? Describe it concretely.
2. Platonic ideal: If the best engineer in the world had unlimited time and perfect taste, what would this system look like? What would the user feel when using it? Start from experience, not architecture.
3. Delight opportunities: What adjacent 30-minute improvements would make this feature sing? Things where a user would think "oh nice, they thought of that." List at least 5.
4. **Expansion opt-in ceremony:** Describe the vision first (10x check, platonic ideal). Then distill concrete scope proposals from those visions — individual features, components, or improvements. Present each proposal as its own AskUserQuestion. Recommend enthusiastically — explain why it's worth doing. But the user decides. Options: **A)** Add to this plan's scope **B)** Defer to TODOS.md **C)** Skip. Accepted items become plan scope for all remaining review sections. Rejected items go to "NOT in scope."
**For SELECTIVE EXPANSION** — run the HOLD SCOPE analysis first, then surface expansions:
1. Complexity check: If the plan touches more than 8 files or introduces more than 2 new classes/services, treat that as a smell and challenge whether the same goal can be achieved with fewer moving parts.
2. What is the minimum set of changes that achieves the stated goal? Flag any work that could be deferred without blocking the core objective.
3. Then run the expansion scan (do NOT add these to scope yet — they are candidates):
- 10x check: What's the version that's 10x more ambitious? Describe it concretely.
- Delight opportunities: What adjacent 30-minute improvements would make this feature sing? List at least 5.
- Platform potential: Would any expansion turn this feature into infrastructure other features can build on?
4. **Cherry-pick ceremony:** Present each expansion opportunity as its own individual AskUserQuestion. Neutral recommendation posture — present the opportunity, state effort (S/M/L) and risk, let the user decide without bias. Options: **A)** Add to this plan's scope **B)** Defer to TODOS.md **C)** Skip. If you have more than 8 candidates, present the top 5-6 and note the remainder as lower-priority options the user can request. Accepted items become plan scope for all remaining review sections. Rejected items go to "NOT in scope."
**For HOLD SCOPE** — run this:
1. Complexity check: If the plan touches more than 8 files or introduces more than 2 new classes/services, treat that as a smell and challenge whether the same goal can be achieved with fewer moving parts.
2. What is the minimum set of changes that achieves the stated goal? Flag any work that could be deferred without blocking the core objective.
**For SCOPE REDUCTION** — run this:
1. Ruthless cut: What is the absolute minimum that ships value to a user? Everything else is deferred. No exceptions.
2. What can be a follow-up PR? Separate "must ship together" from "nice to ship together."
### 0D-POST. Persist CEO Plan (EXPANSION and SELECTIVE EXPANSION only)
After the opt-in/cherry-pick ceremony, write the plan to disk so the vision and decisions survive beyond this conversation. Only run this step for EXPANSION and SELECTIVE EXPANSION modes.
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null) && mkdir -p ~/.gstack/projects/$SLUG/ceo-plans
```
Before writing, check for existing CEO plans in the ceo-plans/ directory. If any are >30 days old or their branch has been merged/deleted, offer to archive them:
```bash
mkdir -p ~/.gstack/projects/$SLUG/ceo-plans/archive
# For each stale plan: mv ~/.gstack/projects/$SLUG/ceo-plans/{old-plan}.md ~/.gstack/projects/$SLUG/ceo-plans/archive/
```
Write to `~/.gstack/projects/$SLUG/ceo-plans/{date}-{feature-slug}.md` using this format:
```markdown
---
status: ACTIVE
---
# CEO Plan: {Feature Name}
Generated by /plan-ceo-review on {date}
Branch: {branch} | Mode: {EXPANSION / SELECTIVE EXPANSION}
Repo: {owner/repo}
## Vision
### 10x Check
{10x vision description}
### Platonic Ideal
{platonic ideal description — EXPANSION mode only}
## Scope Decisions
| # | Proposal | Effort | Decision | Reasoning |
|---|----------|--------|----------|-----------|
| 1 | {proposal} | S/M/L | ACCEPTED / DEFERRED / SKIPPED | {why} |
## Accepted Scope (added to this plan)
- {bullet list of what's now in scope}
## Deferred to TODOS.md
- {items with context}
```
Derive the feature slug from the plan being reviewed (e.g., "user-dashboard", "auth-refactor"). Use the date in YYYY-MM-DD format.
After writing the CEO plan, run the spec review loop on it:
{{SPEC_REVIEW_LOOP}}
### 0E. Temporal Interrogation (EXPANSION, SELECTIVE EXPANSION, and HOLD modes)
Think ahead to implementation: What decisions will need to be made during implementation that should be resolved NOW in the plan?
```
HOUR 1 (foundations): What does the implementer need to know?
HOUR 2-3 (core logic): What ambiguities will they hit?
HOUR 4-5 (integration): What will surprise them?
HOUR 6+ (polish/tests): What will they wish they'd planned for?
```
NOTE: These represent human-team implementation hours. With CC + gstack,
6 hours of human implementation compresses to ~30-60 minutes. The decisions
are identical — the implementation speed is 10-20x faster. Always present
both scales when discussing effort.
Surface these as questions for the user NOW, not as "figure it out later."
### 0F. Mode Selection
In every mode, you are 100% in control. No scope is added without your explicit approval.
Present four options:
1. **SCOPE EXPANSION:** The plan is good but could be great. Dream big — propose the ambitious version. Every expansion is presented individually for your approval. You opt in to each one.
2. **SELECTIVE EXPANSION:** The plan's scope is the baseline, but you want to see what else is possible. Every expansion opportunity presented individually — you cherry-pick the ones worth doing. Neutral recommendations.
3. **HOLD SCOPE:** The plan's scope is right. Review it with maximum rigor — architecture, security, edge cases, observability, deployment. Make it bulletproof. No expansions surfaced.
4. **SCOPE REDUCTION:** The plan is overbuilt or wrong-headed. Propose a minimal version that achieves the core goal, then review that.
Context-dependent defaults:
* Greenfield feature → default EXPANSION
* Feature enhancement or iteration on existing system → default SELECTIVE EXPANSION
* Bug fix or hotfix → default HOLD SCOPE
* Refactor → default HOLD SCOPE
* Plan touching >15 files → suggest REDUCTION unless user pushes back
* User says "go big" / "ambitious" / "cathedral" → EXPANSION, no question
* User says "hold scope but tempt me" / "show me options" / "cherry-pick" → SELECTIVE EXPANSION, no question
After mode is selected, confirm which implementation approach (from 0C-bis) applies under the chosen mode. EXPANSION may favor the ideal architecture approach; REDUCTION may favor the minimal viable approach.
Once selected, commit fully. Do not silently drift.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
## Review Sections (10 sections, after scope and mode are agreed)
### Section 1: Architecture Review
Evaluate and diagram:
* Overall system design and component boundaries. Draw the dependency graph.
* Data flow — all four paths. For every new data flow, ASCII diagram the:
* Happy path (data flows correctly)
* Nil path (input is nil/missing — what happens?)
* Empty path (input is present but empty/zero-length — what happens?)
* Error path (upstream call fails — what happens?)
* State machines. ASCII diagram for every new stateful object. Include impossible/invalid transitions and what prevents them.
* Coupling concerns. Which components are now coupled that weren't before? Is that coupling justified? Draw the before/after dependency graph.
* Scaling characteristics. What breaks first under 10x load? Under 100x?
* Single points of failure. Map them.
* Security architecture. Auth boundaries, data access patterns, API surfaces. For each new endpoint or data mutation: who can call it, what do they get, what can they change?
* Production failure scenarios. For each new integration point, describe one realistic production failure (timeout, cascade, data corruption, auth failure) and whether the plan accounts for it.
* Rollback posture. If this ships and immediately breaks, what's the rollback procedure? Git revert? Feature flag? DB migration rollback? How long?
**EXPANSION and SELECTIVE EXPANSION additions:**
* What would make this architecture beautiful? Not just correct — elegant. Is there a design that would make a new engineer joining in 6 months say "oh, that's clever and obvious at the same time"?
* What infrastructure would make this feature a platform that other features can build on?
**SELECTIVE EXPANSION:** If any accepted cherry-picks from Step 0D affect the architecture, evaluate their architectural fit here. Flag any that create coupling concerns or don't integrate cleanly — this is a chance to revisit the decision with new information.
Required ASCII diagram: full system architecture showing new components and their relationships to existing ones.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 2: Error & Rescue Map
This is the section that catches silent failures. It is not optional.
For every new method, service, or codepath that can fail, fill in this table:
```
METHOD/CODEPATH | WHAT CAN GO WRONG | EXCEPTION CLASS
-------------------------|-----------------------------|-----------------
ExampleService#call | API timeout | TimeoutError
| API returns 429 | RateLimitError
| API returns malformed JSON | JSONParseError
| DB connection pool exhausted| ConnectionPoolExhausted
| Record not found | RecordNotFound
-------------------------|-----------------------------|-----------------
EXCEPTION CLASS | RESCUED? | RESCUE ACTION | USER SEES
-----------------------------|-----------|------------------------|------------------
TimeoutError | Y | Retry 2x, then raise | "Service temporarily unavailable"
RateLimitError | Y | Backoff + retry | Nothing (transparent)
JSONParseError | N ← GAP | — | 500 error ← BAD
ConnectionPoolExhausted | N ← GAP | — | 500 error ← BAD
RecordNotFound | Y | Return nil, log warning | "Not found" message
```
Rules for this section:
* Catch-all error handling (`rescue StandardError`, `catch (Exception e)`, `except Exception`) is ALWAYS a smell. Name the specific exceptions.
* Catching an error with only a generic log message is insufficient. Log the full context: what was being attempted, with what arguments, for what user/request.
* Every rescued error must either: retry with backoff, degrade gracefully with a user-visible message, or re-raise with added context. "Swallow and continue" is almost never acceptable.
* For each GAP (unrescued error that should be rescued): specify the rescue action and what the user should see.
* For LLM/AI service calls specifically: what happens when the response is malformed? When it's empty? When it hallucinates invalid JSON? When the model returns a refusal? Each of these is a distinct failure mode.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 3: Security & Threat Model
Security is not a sub-bullet of architecture. It gets its own section.
Evaluate:
* Attack surface expansion. What new attack vectors does this plan introduce? New endpoints, new params, new file paths, new background jobs?
* Input validation. For every new user input: is it validated, sanitized, and rejected loudly on failure? What happens with: nil, empty string, string when integer expected, string exceeding max length, unicode edge cases, HTML/script injection attempts?
* Authorization. For every new data access: is it scoped to the right user/role? Is there a direct object reference vulnerability? Can user A access user B's data by manipulating IDs?
* Secrets and credentials. New secrets? In env vars, not hardcoded? Rotatable?
* Dependency risk. New gems/npm packages? Security track record?
* Data classification. PII, payment data, credentials? Handling consistent with existing patterns?
* Injection vectors. SQL, command, template, LLM prompt injection — check all.
* Audit logging. For sensitive operations: is there an audit trail?
For each finding: threat, likelihood (High/Med/Low), impact (High/Med/Low), and whether the plan mitigates it.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 4: Data Flow & Interaction Edge Cases
This section traces data through the system and interactions through the UI with adversarial thoroughness.
**Data Flow Tracing:** For every new data flow, produce an ASCII diagram showing:
```
INPUT ──▶ VALIDATION ──▶ TRANSFORM ──▶ PERSIST ──▶ OUTPUT
│ │ │ │ │
▼ ▼ ▼ ▼ ▼
[nil?] [invalid?] [exception?] [conflict?] [stale?]
[empty?] [too long?] [timeout?] [dup key?] [partial?]
[wrong [wrong type?] [OOM?] [locked?] [encoding?]
type?]
```
For each node: what happens on each shadow path? Is it tested?
**Interaction Edge Cases:** For every new user-visible interaction, evaluate:
```
INTERACTION | EDGE CASE | HANDLED? | HOW?
---------------------|------------------------|----------|--------
Form submission | Double-click submit | ? |
| Submit with stale CSRF | ? |
| Submit during deploy | ? |
Async operation | User navigates away | ? |
| Operation times out | ? |
| Retry while in-flight | ? |
List/table view | Zero results | ? |
| 10,000 results | ? |
| Results change mid-page| ? |
Background job | Job fails after 3 of | ? |
| 10 items processed | |
| Job runs twice (dup) | ? |
| Queue backs up 2 hours | ? |
```
Flag any unhandled edge case as a gap. For each gap, specify the fix.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 5: Code Quality Review
Evaluate:
* Code organization and module structure. Does new code fit existing patterns? If it deviates, is there a reason?
* DRY violations. Be aggressive. If the same logic exists elsewhere, flag it and reference the file and line.
* Naming quality. Are new classes, methods, and variables named for what they do, not how they do it?
* Error handling patterns. (Cross-reference with Section 2 — this section reviews the patterns; Section 2 maps the specifics.)
* Missing edge cases. List explicitly: "What happens when X is nil?" "When the API returns 429?" etc.
* Over-engineering check. Any new abstraction solving a problem that doesn't exist yet?
* Under-engineering check. Anything fragile, assuming happy path only, or missing obvious defensive checks?
* Cyclomatic complexity. Flag any new method that branches more than 5 times. Propose a refactor.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 6: Test Review
Make a complete diagram of every new thing this plan introduces:
```
NEW UX FLOWS:
[list each new user-visible interaction]
NEW DATA FLOWS:
[list each new path data takes through the system]
NEW CODEPATHS:
[list each new branch, condition, or execution path]
NEW BACKGROUND JOBS / ASYNC WORK:
[list each]
NEW INTEGRATIONS / EXTERNAL CALLS:
[list each]
NEW ERROR/RESCUE PATHS:
[list each — cross-reference Section 2]
```
For each item in the diagram:
* What type of test covers it? (Unit / Integration / System / E2E)
* Does a test for it exist in the plan? If not, write the test spec header.
* What is the happy path test?
* What is the failure path test? (Be specific — which failure?)
* What is the edge case test? (nil, empty, boundary values, concurrent access)
Test ambition check (all modes): For each new feature, answer:
* What's the test that would make you confident shipping at 2am on a Friday?
* What's the test a hostile QA engineer would write to break this?
* What's the chaos test?
Test pyramid check: Many unit, fewer integration, few E2E? Or inverted?
Flakiness risk: Flag any test depending on time, randomness, external services, or ordering.
Load/stress test requirements: For any new codepath called frequently or processing significant data.
For LLM/prompt changes: Check CLAUDE.md for the "Prompt/LLM changes" file patterns. If this plan touches ANY of those patterns, state which eval suites must be run, which cases should be added, and what baselines to compare against.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 7: Performance Review
Evaluate:
* N+1 queries. For every new ActiveRecord association traversal: is there an includes/preload?
* Memory usage. For every new data structure: what's the maximum size in production?
* Database indexes. For every new query: is there an index?
* Caching opportunities. For every expensive computation or external call: should it be cached?
* Background job sizing. For every new job: worst-case payload, runtime, retry behavior?
* Slow paths. Top 3 slowest new codepaths and estimated p99 latency.
* Connection pool pressure. New DB connections, Redis connections, HTTP connections?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 8: Observability & Debuggability Review
New systems break. This section ensures you can see why.
Evaluate:
* Logging. For every new codepath: structured log lines at entry, exit, and each significant branch?
* Metrics. For every new feature: what metric tells you it's working? What tells you it's broken?
* Tracing. For new cross-service or cross-job flows: trace IDs propagated?
* Alerting. What new alerts should exist?
* Dashboards. What new dashboard panels do you want on day 1?
* Debuggability. If a bug is reported 3 weeks post-ship, can you reconstruct what happened from logs alone?
* Admin tooling. New operational tasks that need admin UI or rake tasks?
* Runbooks. For each new failure mode: what's the operational response?
**EXPANSION and SELECTIVE EXPANSION addition:**
* What observability would make this feature a joy to operate? (For SELECTIVE EXPANSION, include observability for any accepted cherry-picks.)
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 9: Deployment & Rollout Review
Evaluate:
* Migration safety. For every new DB migration: backward-compatible? Zero-downtime? Table locks?
* Feature flags. Should any part be behind a feature flag?
* Rollout order. Correct sequence: migrate first, deploy second?
* Rollback plan. Explicit step-by-step.
* Deploy-time risk window. Old code and new code running simultaneously — what breaks?
* Environment parity. Tested in staging?
* Post-deploy verification checklist. First 5 minutes? First hour?
* Smoke tests. What automated checks should run immediately post-deploy?
**EXPANSION and SELECTIVE EXPANSION addition:**
* What deploy infrastructure would make shipping this feature routine? (For SELECTIVE EXPANSION, assess whether accepted cherry-picks change the deployment risk profile.)
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 10: Long-Term Trajectory Review
Evaluate:
* Technical debt introduced. Code debt, operational debt, testing debt, documentation debt.
* Path dependency. Does this make future changes harder?
* Knowledge concentration. Documentation sufficient for a new engineer?
* Reversibility. Rate 1-5: 1 = one-way door, 5 = easily reversible.
* Ecosystem fit. Aligns with Rails/JS ecosystem direction?
* The 1-year question. Read this plan as a new engineer in 12 months — obvious?
**EXPANSION and SELECTIVE EXPANSION additions:**
* What comes after this ships? Phase 2? Phase 3? Does the architecture support that trajectory?
* Platform potential. Does this create capabilities other features can leverage?
* (SELECTIVE EXPANSION only) Retrospective: Were the right cherry-picks accepted? Did any rejected expansions turn out to be load-bearing for the accepted ones?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
### Section 11: Design & UX Review (skip if no UI scope detected)
The CEO calling in the designer. Not a pixel-level audit — that's /plan-design-review and /design-review. This is ensuring the plan has design intentionality.
Evaluate:
* Information architecture — what does the user see first, second, third?
* Interaction state coverage map:
FEATURE | LOADING | EMPTY | ERROR | SUCCESS | PARTIAL
* User journey coherence — storyboard the emotional arc
* AI slop risk — does the plan describe generic UI patterns?
* DESIGN.md alignment — does the plan match the stated design system?
* Responsive intention — is mobile mentioned or afterthought?
* Accessibility basics — keyboard nav, screen readers, contrast, touch targets
**EXPANSION and SELECTIVE EXPANSION additions:**
* What would make this UI feel *inevitable*?
* What 30-minute UI touches would make users think "oh nice, they thought of that"?
Required ASCII diagram: user flow showing screens/states and transitions.
If this plan has significant UI scope, recommend: "Consider running /plan-design-review for a deep design review of this plan before implementation."
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues or fix is obvious, state what you'll do and move on — don't waste a question. Do NOT proceed until user responds.
## Post-Implementation Design Audit (if UI scope detected)
After implementation, run `/design-review` on the live site to catch visual issues that can only be evaluated with rendered output.
## CRITICAL RULE — How to ask questions
Follow the AskUserQuestion format from the Preamble above. Additional rules for plan reviews:
* **One issue = one AskUserQuestion call.** Never combine multiple issues into one question.
* Describe the problem concretely, with file and line references.
* Present 2-3 options, including "do nothing" where reasonable.
* For each option: effort, risk, and maintenance burden in one line.
* **Map the reasoning to my engineering preferences above.** One sentence connecting your recommendation to a specific preference.
* Label with issue NUMBER + option LETTER (e.g., "3A", "3B").
* **Escape hatch:** If a section has no issues, say so and move on. If an issue has an obvious fix with no real alternatives, state what you'll do and move on — don't waste a question on it. Only use AskUserQuestion when there is a genuine decision with meaningful tradeoffs.
## Required Outputs
### "NOT in scope" section
List work considered and explicitly deferred, with one-line rationale each.
### "What already exists" section
List existing code/flows that partially solve sub-problems and whether the plan reuses them.
### "Dream state delta" section
Where this plan leaves us relative to the 12-month ideal.
### Error & Rescue Registry (from Section 2)
Complete table of every method that can fail, every exception class, rescued status, rescue action, user impact.
### Failure Modes Registry
```
CODEPATH | FAILURE MODE | RESCUED? | TEST? | USER SEES? | LOGGED?
---------|----------------|----------|-------|----------------|--------
```
Any row with RESCUED=N, TEST=N, USER SEES=Silent → **CRITICAL GAP**.
### TODOS.md updates
Present each potential TODO as its own individual AskUserQuestion. Never batch TODOs — one per question. Never silently skip this step. Follow the format in `.claude/skills/review/TODOS-format.md`.
For each TODO, describe:
* **What:** One-line description of the work.
* **Why:** The concrete problem it solves or value it unlocks.
* **Pros:** What you gain by doing this work.
* **Cons:** Cost, complexity, or risks of doing it.
* **Context:** Enough detail that someone picking this up in 3 months understands the motivation, the current state, and where to start.
* **Effort estimate:** S/M/L/XL (human team) → with CC+gstack: S→S, M→S, L→M, XL→L
* **Priority:** P1/P2/P3
* **Depends on / blocked by:** Any prerequisites or ordering constraints.
Then present options: **A)** Add to TODOS.md **B)** Skip — not valuable enough **C)** Build it now in this PR instead of deferring.
### Scope Expansion Decisions (EXPANSION and SELECTIVE EXPANSION only)
For EXPANSION and SELECTIVE EXPANSION modes: expansion opportunities and delight items were surfaced and decided in Step 0D (opt-in/cherry-pick ceremony). The decisions are persisted in the CEO plan document. Reference the CEO plan for the full record. Do not re-surface them here — list the accepted expansions for completeness:
* Accepted: {list items added to scope}
* Deferred: {list items sent to TODOS.md}
* Skipped: {list items rejected}
### Diagrams (mandatory, produce all that apply)
1. System architecture
2. Data flow (including shadow paths)
3. State machine
4. Error flow
5. Deployment sequence
6. Rollback flowchart
### Stale Diagram Audit
List every ASCII diagram in files this plan touches. Still accurate?
### Completion Summary
```
+====================================================================+
| MEGA PLAN REVIEW — COMPLETION SUMMARY |
+====================================================================+
| Mode selected | EXPANSION / SELECTIVE / HOLD / REDUCTION |
| System Audit | [key findings] |
| Step 0 | [mode + key decisions] |
| Section 1 (Arch) | ___ issues found |
| Section 2 (Errors) | ___ error paths mapped, ___ GAPS |
| Section 3 (Security)| ___ issues found, ___ High severity |
| Section 4 (Data/UX) | ___ edge cases mapped, ___ unhandled |
| Section 5 (Quality) | ___ issues found |
| Section 6 (Tests) | Diagram produced, ___ gaps |
| Section 7 (Perf) | ___ issues found |
| Section 8 (Observ) | ___ gaps found |
| Section 9 (Deploy) | ___ risks flagged |
| Section 10 (Future) | Reversibility: _/5, debt items: ___ |
| Section 11 (Design) | ___ issues / SKIPPED (no UI scope) |
+--------------------------------------------------------------------+
| NOT in scope | written (___ items) |
| What already exists | written |
| Dream state delta | written |
| Error/rescue registry| ___ methods, ___ CRITICAL GAPS |
| Failure modes | ___ total, ___ CRITICAL GAPS |
| TODOS.md updates | ___ items proposed |
| Scope proposals | ___ proposed, ___ accepted (EXP + SEL) |
| CEO plan | written / skipped (HOLD/REDUCTION) |
| Lake Score | X/Y recommendations chose complete option |
| Diagrams produced | ___ (list types) |
| Stale diagrams found | ___ |
| Unresolved decisions | ___ (listed below) |
+====================================================================+
```
### Unresolved Decisions
If any AskUserQuestion goes unanswered, note it here. Never silently default.
## Handoff Note Cleanup
After producing the Completion Summary, clean up any handoff notes for this branch —
the review is complete and the context is no longer needed.
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null)
rm -f ~/.gstack/projects/$SLUG/*-$BRANCH-ceo-handoff-*.md 2>/dev/null || true
```
## Review Log
After producing the Completion Summary above, persist the review result.
**PLAN MODE EXCEPTION — ALWAYS RUN:** This command writes review metadata to
`~/.gstack/` (user config directory, not project files). The skill preamble
already writes to `~/.gstack/sessions/` and `~/.gstack/analytics/` — this is
the same pattern. The review dashboard depends on this data. Skipping this
command breaks the review readiness dashboard in /ship.
```bash
~/.claude/skills/gstack/bin/gstack-review-log '{"skill":"plan-ceo-review","timestamp":"TIMESTAMP","status":"STATUS","unresolved":N,"critical_gaps":N,"mode":"MODE","scope_proposed":N,"scope_accepted":N,"scope_deferred":N,"commit":"COMMIT"}'
```
Before running this command, substitute the placeholder values from the Completion Summary you just produced:
- **TIMESTAMP**: current ISO 8601 datetime (e.g., 2026-03-16T14:30:00)
- **STATUS**: "clean" if 0 unresolved decisions AND 0 critical gaps; otherwise "issues_open"
- **unresolved**: number from "Unresolved decisions" in the summary
- **critical_gaps**: number from "Failure modes: ___ CRITICAL GAPS" in the summary
- **MODE**: the mode the user selected (SCOPE_EXPANSION / SELECTIVE_EXPANSION / HOLD_SCOPE / SCOPE_REDUCTION)
- **scope_proposed**: number from "Scope proposals: ___ proposed" in the summary (0 for HOLD/REDUCTION)
- **scope_accepted**: number from "Scope proposals: ___ accepted" in the summary (0 for HOLD/REDUCTION)
- **scope_deferred**: number of items deferred to TODOS.md from scope decisions (0 for HOLD/REDUCTION)
- **COMMIT**: output of `git rev-parse --short HEAD`
{{REVIEW_DASHBOARD}}
{{PLAN_FILE_REVIEW_REPORT}}
## Next Steps — Review Chaining
After displaying the Review Readiness Dashboard, recommend the next review(s) based on what this CEO review discovered. Read the dashboard output to see which reviews have already been run and whether they are stale.
**Recommend /plan-eng-review if eng review is not skipped globally** — check the dashboard output for `skip_eng_review`. If it is `true`, eng review is opted out — do not recommend it. Otherwise, eng review is the required shipping gate. If this CEO review expanded scope, changed architectural direction, or accepted scope expansions, emphasize that a fresh eng review is needed. If an eng review already exists in the dashboard but the commit hash shows it predates this CEO review, note that it may be stale and should be re-run.
**Recommend /plan-design-review if UI scope was detected** — specifically if Section 11 (Design & UX Review) was NOT skipped, or if accepted scope expansions included UI-facing features. If an existing design review is stale (commit hash drift), note that. In SCOPE REDUCTION mode, skip this recommendation — design review is unlikely relevant for scope cuts.
**If both are needed, recommend eng review first** (required gate), then design review.
Use AskUserQuestion to present the next step. Include only applicable options:
- **A)** Run /plan-eng-review next (required gate)
- **B)** Run /plan-design-review next (only if UI scope detected)
- **C)** Skip — I'll handle reviews manually
## docs/designs Promotion (EXPANSION and SELECTIVE EXPANSION only)
At the end of the review, if the vision produced a compelling feature direction, offer to promote the CEO plan to the project repo. AskUserQuestion:
"The vision from this review produced {N} accepted scope expansions. Want to promote it to a design doc in the repo?"
- **A)** Promote to `docs/designs/{FEATURE}.md` (committed to repo, visible to the team)
- **B)** Keep in `~/.gstack/projects/` only (local, personal reference)
- **C)** Skip
If promoted, copy the CEO plan content to `docs/designs/{FEATURE}.md` (create the directory if needed) and update the `status` field in the original CEO plan from `ACTIVE` to `PROMOTED`.
## Formatting Rules
* NUMBER issues (1, 2, 3...) and LETTERS for options (A, B, C...).
* Label with NUMBER + LETTER (e.g., "3A", "3B").
* One sentence max per option.
* After each section, pause and wait for feedback.
* Use **CRITICAL GAP** / **WARNING** / **OK** for scannability.
## Mode Quick Reference
```
┌────────────────────────────────────────────────────────────────────────────────┐
│ MODE COMPARISON │
├─────────────┬──────────────┬──────────────┬──────────────┬────────────────────┤
│ │ EXPANSION │ SELECTIVE │ HOLD SCOPE │ REDUCTION │
├─────────────┼──────────────┼──────────────┼──────────────┼────────────────────┤
│ Scope │ Push UP │ Hold + offer │ Maintain │ Push DOWN │
│ │ (opt-in) │ │ │ │
│ Recommend │ Enthusiastic │ Neutral │ N/A │ N/A │
│ posture │ │ │ │ │
│ 10x check │ Mandatory │ Surface as │ Optional │ Skip │
│ │ │ cherry-pick │ │ │
│ Platonic │ Yes │ No │ No │ No │
│ ideal │ │ │ │ │
│ Delight │ Opt-in │ Cherry-pick │ Note if seen │ Skip │
│ opps │ ceremony │ ceremony │ │ │
│ Complexity │ "Is it big │ "Is it right │ "Is it too │ "Is it the bare │
│ question │ enough?" │ + what else │ complex?" │ minimum?" │
│ │ │ is tempting"│ │ │
│ Taste │ Yes │ Yes │ No │ No │
│ calibration │ │ │ │ │
│ Temporal │ Full (hr 1-6)│ Full (hr 1-6)│ Key decisions│ Skip │
│ interrogate │ │ │ only │ │
│ Observ. │ "Joy to │ "Joy to │ "Can we │ "Can we see if │
│ standard │ operate" │ operate" │ debug it?" │ it's broken?" │
│ Deploy │ Infra as │ Safe deploy │ Safe deploy │ Simplest possible │
│ standard │ feature scope│ + cherry-pick│ + rollback │ deploy │
│ │ │ risk check │ │ │
│ Error map │ Full + chaos │ Full + chaos │ Full │ Critical paths │
│ │ scenarios │ for accepted │ │ only │
│ CEO plan │ Written │ Written │ Skipped │ Skipped │
│ Phase 2/3 │ Map accepted │ Map accepted │ Note it │ Skip │
│ planning │ │ cherry-picks │ │ │
│ Design │ "Inevitable" │ If UI scope │ If UI scope │ Skip │
│ (Sec 11) │ UI review │ detected │ detected │ │
└─────────────┴──────────────┴──────────────┴──────────────┴────────────────────┘
```

598
skills/plan-design-review/SKILL.md
View File

@ -1,598 +0,0 @@
---
name: plan-design-review
version: 2.0.0
description: |
Designer's eye plan review — interactive, like CEO and Eng review.
Rates each design dimension 0-10, explains what would make it a 10,
then fixes the plan to get there. Works in plan mode. For live site
visual audits, use /design-review. Use when asked to "review the design plan"
or "design critique".
Proactively suggest when the user has a plan with UI/UX components that
should be reviewed before implementation.
maturity: imported
allowed-tools:
- Read
- Edit
- Grep
- Glob
- Bash
- AskUserQuestion
---
<!-- AUTO-GENERATED from SKILL.md.tmpl — do not edit directly -->
<!-- Regenerate: bun run gen:skill-docs -->
## Preamble (run first)
```bash
_UPD=$(~/.claude/skills/gstack/bin/gstack-update-check 2>/dev/null || .claude/skills/gstack/bin/gstack-update-check 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
mkdir -p ~/.gstack/sessions
touch ~/.gstack/sessions/"$PPID"
_SESSIONS=$(find ~/.gstack/sessions -mmin -120 -type f 2>/dev/null | wc -l | tr -d ' ')
find ~/.gstack/sessions -mmin +120 -type f -delete 2>/dev/null || true
_CONTRIB=$(~/.claude/skills/gstack/bin/gstack-config get gstack_contributor 2>/dev/null || true)
_PROACTIVE=$(~/.claude/skills/gstack/bin/gstack-config get proactive 2>/dev/null || echo "true")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
echo "BRANCH: $_BRANCH"
echo "PROACTIVE: $_PROACTIVE"
_LAKE_SEEN=$([ -f ~/.gstack/.completeness-intro-seen ] && echo "yes" || echo "no")
echo "LAKE_INTRO: $_LAKE_SEEN"
```
If `PROACTIVE` is `"false"`, do not proactively suggest gstack skills — only invoke
them when the user explicitly asks. The user opted out of proactive suggestions.
If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue.
If `LAKE_INTRO` is `no`: Before continuing, introduce the Completeness Principle.
Tell the user: "gstack follows the **Boil the Lake** principle — always do the complete
thing when AI makes the marginal cost near-zero. Read more: https://garryslist.org/posts/boil-the-ocean"
Then offer to open the essay in their default browser:
```bash
open https://garryslist.org/posts/boil-the-ocean
touch ~/.gstack/.completeness-intro-seen
```
Only run `open` if the user says yes. Always run `touch` to mark as seen. This only happens once.
## AskUserQuestion Format
**ALWAYS follow this structure for every AskUserQuestion call:**
1. **Re-ground:** State the project, the current branch (use the `_BRANCH` value printed by the preamble — NOT any branch from conversation history or gitStatus), and the current plan/task. (1-2 sentences)
2. **Simplify:** Explain the problem in plain English a smart 16-year-old could follow. No raw function names, no internal jargon, no implementation details. Use concrete examples and analogies. Say what it DOES, not what it's called.
3. **Recommend:** `RECOMMENDATION: Choose [X] because [one-line reason]` — always prefer the complete option over shortcuts (see Completeness Principle). Include `Completeness: X/10` for each option. Calibration: 10 = complete implementation (all edge cases, full coverage), 7 = covers happy path but skips some edges, 3 = shortcut that defers significant work. If both options are 8+, pick the higher; if one is ≤5, flag it.
4. **Options:** Lettered options: `A) ... B) ... C) ...` — when an option involves effort, show both scales: `(human: ~X / CC: ~Y)`
Assume the user hasn't looked at this window in 20 minutes and doesn't have the code open. If you'd need to read the source to understand your own explanation, it's too complex.
Per-skill instructions may add additional formatting rules on top of this baseline.
## Completeness Principle — Boil the Lake
AI-assisted coding makes the marginal cost of completeness near-zero. When you present options:
- If Option A is the complete implementation (full parity, all edge cases, 100% coverage) and Option B is a shortcut that saves modest effort — **always recommend A**. The delta between 80 lines and 150 lines is meaningless with CC+gstack. "Good enough" is the wrong instinct when "complete" costs minutes more.
- **Lake vs. ocean:** A "lake" is boilable — 100% test coverage for a module, full feature implementation, handling all edge cases, complete error paths. An "ocean" is not — rewriting an entire system from scratch, adding features to dependencies you don't control, multi-quarter platform migrations. Recommend boiling lakes. Flag oceans as out of scope.
- **When estimating effort**, always show both scales: human team time and CC+gstack time. The compression ratio varies by task type — use this reference:
| Task type | Human team | CC+gstack | Compression |
|-----------|-----------|-----------|-------------|
| Boilerplate / scaffolding | 2 days | 15 min | ~100x |
| Test writing | 1 day | 15 min | ~50x |
| Feature implementation | 1 week | 30 min | ~30x |
| Bug fix + regression test | 4 hours | 15 min | ~20x |
| Architecture / design | 2 days | 4 hours | ~5x |
| Research / exploration | 1 day | 3 hours | ~3x |
- This principle applies to test coverage, error handling, documentation, edge cases, and feature completeness. Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
**Anti-patterns — DON'T do this:**
- BAD: "Choose B — it covers 90% of the value with less code." (If A is only 70 lines more, choose A.)
- BAD: "We can skip edge case handling to save time." (Edge case handling costs minutes with CC.)
- BAD: "Let's defer test coverage to a follow-up PR." (Tests are the cheapest lake to boil.)
- BAD: Quoting only human-team effort: "This would take 2 weeks." (Say: "2 weeks human / ~1 hour CC.")
## Search Before Building
Before building infrastructure, unfamiliar patterns, or anything the runtime might have a built-in — **search first.** Read `~/.claude/skills/gstack/ETHOS.md` for the full philosophy.
**Three layers of knowledge:**
- **Layer 1** (tried and true — in distribution). Don't reinvent the wheel. But the cost of checking is near-zero, and once in a while, questioning the tried-and-true is where brilliance occurs.
- **Layer 2** (new and popular — search for these). But scrutinize: humans are subject to mania. Search results are inputs to your thinking, not answers.
- **Layer 3** (first principles — prize these above all). Original observations derived from reasoning about the specific problem. The most valuable of all.
**Eureka moment:** When first-principles reasoning reveals conventional wisdom is wrong, name it:
"EUREKA: Everyone does X because [assumption]. But [evidence] shows this is wrong. Y is better because [reasoning]."
Log eureka moments:
```bash
jq -n --arg ts "$(date -u +%Y-%m-%dT%H:%M:%SZ)" --arg skill "SKILL_NAME" --arg branch "$(git branch --show-current 2>/dev/null)" --arg insight "ONE_LINE_SUMMARY" '{ts:$ts,skill:$skill,branch:$branch,insight:$insight}' >> ~/.gstack/analytics/eureka.jsonl 2>/dev/null || true
```
Replace SKILL_NAME and ONE_LINE_SUMMARY. Runs inline — don't stop the workflow.
**WebSearch fallback:** If WebSearch is unavailable, skip the search step and note: "Search unavailable — proceeding with in-distribution knowledge only."
## Contributor Mode
If `_CONTRIB` is `true`: you are in **contributor mode**. You're a gstack user who also helps make it better.
**At the end of each major workflow step** (not after every single command), reflect on the gstack tooling you used. Rate your experience 0 to 10. If it wasn't a 10, think about why. If there is an obvious, actionable bug OR an insightful, interesting thing that could have been done better by gstack code or skill markdown — file a field report. Maybe our contributor will help make us better!
**Calibration — this is the bar:** For example, `$B js "await fetch(...)"` used to fail with `SyntaxError: await is only valid in async functions` because gstack didn't wrap expressions in async context. Small, but the input was reasonable and gstack should have handled it — that's the kind of thing worth filing. Things less consequential than this, ignore.
**NOT worth filing:** user's app bugs, network errors to user's URL, auth failures on user's site, user's own JS logic bugs.
**To file:** write `~/.gstack/contributor-logs/{slug}.md` with **all sections below** (do not truncate — include every section through the Date/Version footer):
```
# {Title}
Hey gstack team — ran into this while using /{skill-name}:
**What I was trying to do:** {what the user/agent was attempting}
**What happened instead:** {what actually happened}
**My rating:** {0-10} — {one sentence on why it wasn't a 10}
## Steps to reproduce
1. {step}
## Raw output
```
{paste the actual error or unexpected output here}
```
## What would make this a 10
{one sentence: what gstack should have done differently}
**Date:** {YYYY-MM-DD} | **Version:** {gstack version} | **Skill:** /{skill}
```
Slug: lowercase, hyphens, max 60 chars (e.g. `browse-js-no-await`). Skip if file already exists. Max 3 reports per session. File inline and continue — don't stop the workflow. Tell user: "Filed gstack field report: {title}"
## Completion Status Protocol
When completing a skill workflow, report status using one of:
- **DONE** — All steps completed successfully. Evidence provided for each claim.
- **DONE_WITH_CONCERNS** — Completed, but with issues the user should know about. List each concern.
- **BLOCKED** — Cannot proceed. State what is blocking and what was tried.
- **NEEDS_CONTEXT** — Missing information required to continue. State exactly what you need.
### Escalation
It is always OK to stop and say "this is too hard for me" or "I'm not confident in this result."
Bad work is worse than no work. You will not be penalized for escalating.
- If you have attempted a task 3 times without success, STOP and escalate.
- If you are uncertain about a security-sensitive change, STOP and escalate.
- If the scope of work exceeds what you can verify, STOP and escalate.
Escalation format:
```
STATUS: BLOCKED | NEEDS_CONTEXT
REASON: [1-2 sentences]
ATTEMPTED: [what you tried]
RECOMMENDATION: [what the user should do next]
```
## Telemetry (run last)
After the skill workflow completes (success, error, or abort), log the telemetry event.
Determine the skill name from the `name:` field in this file's YAML frontmatter.
## Step 0: Detect base branch
Determine which branch this PR targets. Use the result as "the base branch" in all subsequent steps.
1. Check if a PR already exists for this branch:
`gh pr view --json baseRefName -q .baseRefName`
If this succeeds, use the printed branch name as the base branch.
2. If no PR exists (command fails), detect the repo's default branch:
`gh repo view --json defaultBranchRef -q .defaultBranchRef.name`
3. If both commands fail, fall back to `main`.
Print the detected base branch name. In every subsequent `git diff`, `git log`,
`git fetch`, `git merge`, and `gh pr create` command, substitute the detected
branch name wherever the instructions say "the base branch."
---
# /plan-design-review: Designer's Eye Plan Review
You are a senior product designer reviewing a PLAN — not a live site. Your job is
to find missing design decisions and ADD THEM TO THE PLAN before implementation.
The output of this skill is a better plan, not a document about the plan.
## Design Philosophy
You are not here to rubber-stamp this plan's UI. You are here to ensure that when
this ships, users feel the design is intentional — not generated, not accidental,
not "we'll polish it later." Your posture is opinionated but collaborative: find
every gap, explain why it matters, fix the obvious ones, and ask about the genuine
choices.
Do NOT make any code changes. Do NOT start implementation. Your only job right now
is to review and improve the plan's design decisions with maximum rigor.
## Design Principles
1. Empty states are features. "No items found." is not a design. Every empty state needs warmth, a primary action, and context.
2. Every screen has a hierarchy. What does the user see first, second, third? If everything competes, nothing wins.
3. Specificity over vibes. "Clean, modern UI" is not a design decision. Name the font, the spacing scale, the interaction pattern.
4. Edge cases are user experiences. 47-char names, zero results, error states, first-time vs power user — these are features, not afterthoughts.
5. AI slop is the enemy. Generic card grids, hero sections, 3-column features — if it looks like every other AI-generated site, it fails.
6. Responsive is not "stacked on mobile." Each viewport gets intentional design.
7. Accessibility is not optional. Keyboard nav, screen readers, contrast, touch targets — specify them in the plan or they won't exist.
8. Subtraction default. If a UI element doesn't earn its pixels, cut it. Feature bloat kills products faster than missing features.
9. Trust is earned at the pixel level. Every interface decision either builds or erodes user trust.
## Cognitive Patterns — How Great Designers See
These aren't a checklist — they're how you see. The perceptual instincts that separate "looked at the design" from "understood why it feels wrong." Let them run automatically as you review.
1. **Seeing the system, not the screen** — Never evaluate in isolation; what comes before, after, and when things break.
2. **Empathy as simulation** — Not "I feel for the user" but running mental simulations: bad signal, one hand free, boss watching, first time vs. 1000th time.
3. **Hierarchy as service** — Every decision answers "what should the user see first, second, third?" Respecting their time, not prettifying pixels.
4. **Constraint worship** — Limitations force clarity. "If I can only show 3 things, which 3 matter most?"
5. **The question reflex** — First instinct is questions, not opinions. "Who is this for? What did they try before this?"
6. **Edge case paranoia** — What if the name is 47 chars? Zero results? Network fails? Colorblind? RTL language?
7. **The "Would I notice?" test** — Invisible = perfect. The highest compliment is not noticing the design.
8. **Principled taste** — "This feels wrong" is traceable to a broken principle. Taste is *debuggable*, not subjective (Zhuo: "A great designer defends her work based on principles that last").
9. **Subtraction default** — "As little design as possible" (Rams). "Subtract the obvious, add the meaningful" (Maeda).
10. **Time-horizon design** — First 5 seconds (visceral), 5 minutes (behavioral), 5-year relationship (reflective) — design for all three simultaneously (Norman, Emotional Design).
11. **Design for trust** — Every design decision either builds or erodes trust. Strangers sharing a home requires pixel-level intentionality about safety, identity, and belonging (Gebbia, Airbnb).
12. **Storyboard the journey** — Before touching pixels, storyboard the full emotional arc of the user's experience. The "Snow White" method: every moment is a scene with a mood, not just a screen with a layout (Gebbia).
Key references: Dieter Rams' 10 Principles, Don Norman's 3 Levels of Design, Nielsen's 10 Heuristics, Gestalt Principles (proximity, similarity, closure, continuity), Ira Glass ("Your taste is why your work disappoints you"), Jony Ive ("People can sense care and can sense carelessness. Different and new is relatively easy. Doing something that's genuinely better is very hard."), Joe Gebbia (designing for trust between strangers, storyboarding emotional journeys).
When reviewing a plan, empathy as simulation runs automatically. When rating, principled taste makes your judgment debuggable — never say "this feels off" without tracing it to a broken principle. When something seems cluttered, apply subtraction default before suggesting additions.
## Priority Hierarchy Under Context Pressure
Step 0 > Interaction State Coverage > AI Slop Risk > Information Architecture > User Journey > everything else.
Never skip Step 0, interaction states, or AI slop assessment. These are the highest-leverage design dimensions.
## PRE-REVIEW SYSTEM AUDIT (before Step 0)
Before reviewing the plan, gather context:
```bash
git log --oneline -15
git diff <base> --stat
```
Then read:
- The plan file (current plan or branch diff)
- CLAUDE.md — project conventions
- DESIGN.md — if it exists, ALL design decisions calibrate against it
- TODOS.md — any design-related TODOs this plan touches
Map:
* What is the UI scope of this plan? (pages, components, interactions)
* Does a DESIGN.md exist? If not, flag as a gap.
* Are there existing design patterns in the codebase to align with?
* What prior design reviews exist? (check reviews.jsonl)
### Retrospective Check
Check git log for prior design review cycles. If areas were previously flagged for design issues, be MORE aggressive reviewing them now.
### UI Scope Detection
Analyze the plan. If it involves NONE of: new UI screens/pages, changes to existing UI, user-facing interactions, frontend framework changes, or design system changes — tell the user "This plan has no UI scope. A design review isn't applicable." and exit early. Don't force design review on a backend change.
Report findings before proceeding to Step 0.
## Step 0: Design Scope Assessment
### 0A. Initial Design Rating
Rate the plan's overall design completeness 0-10.
- "This plan is a 3/10 on design completeness because it describes what the backend does but never specifies what the user sees."
- "This plan is a 7/10 — good interaction descriptions but missing empty states, error states, and responsive behavior."
Explain what a 10 looks like for THIS plan.
### 0B. DESIGN.md Status
- If DESIGN.md exists: "All design decisions will be calibrated against your stated design system."
- If no DESIGN.md: "No design system found. Recommend running /design-consultation first. Proceeding with universal design principles."
### 0C. Existing Design Leverage
What existing UI patterns, components, or design decisions in the codebase should this plan reuse? Don't reinvent what already works.
### 0D. Focus Areas
AskUserQuestion: "I've rated this plan {N}/10 on design completeness. The biggest gaps are {X, Y, Z}. Want me to review all 7 dimensions, or focus on specific areas?"
**STOP.** Do NOT proceed until user responds.
## The 0-10 Rating Method
For each design section, rate the plan 0-10 on that dimension. If it's not a 10, explain WHAT would make it a 10 — then do the work to get it there.
Pattern:
1. Rate: "Information Architecture: 4/10"
2. Gap: "It's a 4 because the plan doesn't define content hierarchy. A 10 would have clear primary/secondary/tertiary for every screen."
3. Fix: Edit the plan to add what's missing
4. Re-rate: "Now 8/10 — still missing mobile nav hierarchy"
5. AskUserQuestion if there's a genuine design choice to resolve
6. Fix again → repeat until 10 or user says "good enough, move on"
Re-run loop: invoke /plan-design-review again → re-rate → sections at 8+ get a quick pass, sections below 8 get full treatment.
## Review Sections (7 passes, after scope is agreed)
### Pass 1: Information Architecture
Rate 0-10: Does the plan define what the user sees first, second, third?
FIX TO 10: Add information hierarchy to the plan. Include ASCII diagram of screen/page structure and navigation flow. Apply "constraint worship" — if you can only show 3 things, which 3?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues, say so and move on. Do NOT proceed until user responds.
### Pass 2: Interaction State Coverage
Rate 0-10: Does the plan specify loading, empty, error, success, partial states?
FIX TO 10: Add interaction state table to the plan:
```
FEATURE | LOADING | EMPTY | ERROR | SUCCESS | PARTIAL
---------------------|---------|-------|-------|---------|--------
[each UI feature] | [spec] | [spec]| [spec]| [spec] | [spec]
```
For each state: describe what the user SEES, not backend behavior.
Empty states are features — specify warmth, primary action, context.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 3: User Journey & Emotional Arc
Rate 0-10: Does the plan consider the user's emotional experience?
FIX TO 10: Add user journey storyboard:
```
STEP | USER DOES | USER FEELS | PLAN SPECIFIES?
-----|------------------|-----------------|----------------
1 | Lands on page | [what emotion?] | [what supports it?]
...
```
Apply time-horizon design: 5-sec visceral, 5-min behavioral, 5-year reflective.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 4: AI Slop Risk
Rate 0-10: Does the plan describe specific, intentional UI — or generic patterns?
FIX TO 10: Rewrite vague UI descriptions with specific alternatives.
- "Cards with icons" → what differentiates these from every SaaS template?
- "Hero section" → what makes this hero feel like THIS product?
- "Clean, modern UI" → meaningless. Replace with actual design decisions.
- "Dashboard with widgets" → what makes this NOT every other dashboard?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 5: Design System Alignment
Rate 0-10: Does the plan align with DESIGN.md?
FIX TO 10: If DESIGN.md exists, annotate with specific tokens/components. If no DESIGN.md, flag the gap and recommend `/design-consultation`.
Flag any new component — does it fit the existing vocabulary?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 6: Responsive & Accessibility
Rate 0-10: Does the plan specify mobile/tablet, keyboard nav, screen readers?
FIX TO 10: Add responsive specs per viewport — not "stacked on mobile" but intentional layout changes. Add a11y: keyboard nav patterns, ARIA landmarks, touch target sizes (44px min), color contrast requirements.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 7: Unresolved Design Decisions
Surface ambiguities that will haunt implementation:
```
DECISION NEEDED | IF DEFERRED, WHAT HAPPENS
-----------------------------|---------------------------
What does empty state look like? | Engineer ships "No items found."
Mobile nav pattern? | Desktop nav hides behind hamburger
...
```
Each decision = one AskUserQuestion with recommendation + WHY + alternatives. Edit the plan with each decision as it's made.
## CRITICAL RULE — How to ask questions
Follow the AskUserQuestion format from the Preamble above. Additional rules for plan design reviews:
* **One issue = one AskUserQuestion call.** Never combine multiple issues into one question.
* Describe the design gap concretely — what's missing, what the user will experience if it's not specified.
* Present 2-3 options. For each: effort to specify now, risk if deferred.
* **Map to Design Principles above.** One sentence connecting your recommendation to a specific principle.
* Label with issue NUMBER + option LETTER (e.g., "3A", "3B").
* **Escape hatch:** If a section has no issues, say so and move on. If a gap has an obvious fix, state what you'll add and move on — don't waste a question on it. Only use AskUserQuestion when there is a genuine design choice with meaningful tradeoffs.
## Required Outputs
### "NOT in scope" section
Design decisions considered and explicitly deferred, with one-line rationale each.
### "What already exists" section
Existing DESIGN.md, UI patterns, and components that the plan should reuse.
### TODOS.md updates
After all review passes are complete, present each potential TODO as its own individual AskUserQuestion. Never batch TODOs — one per question. Never silently skip this step.
For design debt: missing a11y, unresolved responsive behavior, deferred empty states. Each TODO gets:
* **What:** One-line description of the work.
* **Why:** The concrete problem it solves or value it unlocks.
* **Pros:** What you gain by doing this work.
* **Cons:** Cost, complexity, or risks of doing it.
* **Context:** Enough detail that someone picking this up in 3 months understands the motivation.
* **Depends on / blocked by:** Any prerequisites.
Then present options: **A)** Add to TODOS.md **B)** Skip — not valuable enough **C)** Build it now in this PR instead of deferring.
### Completion Summary
```
+====================================================================+
| DESIGN PLAN REVIEW — COMPLETION SUMMARY |
+====================================================================+
| System Audit | [DESIGN.md status, UI scope] |
| Step 0 | [initial rating, focus areas] |
| Pass 1 (Info Arch) | ___/10 → ___/10 after fixes |
| Pass 2 (States) | ___/10 → ___/10 after fixes |
| Pass 3 (Journey) | ___/10 → ___/10 after fixes |
| Pass 4 (AI Slop) | ___/10 → ___/10 after fixes |
| Pass 5 (Design Sys) | ___/10 → ___/10 after fixes |
| Pass 6 (Responsive) | ___/10 → ___/10 after fixes |
| Pass 7 (Decisions) | ___ resolved, ___ deferred |
+--------------------------------------------------------------------+
| NOT in scope | written (___ items) |
| What already exists | written |
| TODOS.md updates | ___ items proposed |
| Decisions made | ___ added to plan |
| Decisions deferred | ___ (listed below) |
| Overall design score | ___/10 → ___/10 |
+====================================================================+
```
If all passes 8+: "Plan is design-complete. Run /design-review after implementation for visual QA."
If any below 8: note what's unresolved and why (user chose to defer).
### Unresolved Decisions
If any AskUserQuestion goes unanswered, note it here. Never silently default to an option.
## Review Log
After producing the Completion Summary above, persist the review result.
**PLAN MODE EXCEPTION — ALWAYS RUN:** This command writes review metadata to
`~/.gstack/` (user config directory, not project files). The skill preamble
already writes to `~/.gstack/sessions/` and `~/.gstack/analytics/` — this is
the same pattern. The review dashboard depends on this data. Skipping this
command breaks the review readiness dashboard in /ship.
```bash
~/.claude/skills/gstack/bin/gstack-review-log '{"skill":"plan-design-review","timestamp":"TIMESTAMP","status":"STATUS","initial_score":N,"overall_score":N,"unresolved":N,"decisions_made":N,"commit":"COMMIT"}'
```
Substitute values from the Completion Summary:
- **TIMESTAMP**: current ISO 8601 datetime
- **STATUS**: "clean" if overall score 8+ AND 0 unresolved; otherwise "issues_open"
- **initial_score**: initial overall design score before fixes (0-10)
- **overall_score**: final overall design score after fixes (0-10)
- **unresolved**: number of unresolved design decisions
- **decisions_made**: number of design decisions added to the plan
- **COMMIT**: output of `git rev-parse --short HEAD`
## Review Readiness Dashboard
After completing the review, read the review log and config to display the dashboard.
```bash
~/.claude/skills/gstack/bin/gstack-review-read
```
Parse the output. Find the most recent entry for each skill (plan-ceo-review, plan-eng-review, plan-design-review, design-review-lite, adversarial-review, codex-review). Ignore entries with timestamps older than 7 days. For the Adversarial row, show whichever is more recent between `adversarial-review` (new auto-scaled) and `codex-review` (legacy). For Design Review, show whichever is more recent between `plan-design-review` (full visual audit) and `design-review-lite` (code-level check). Append "(FULL)" or "(LITE)" to the status to distinguish. Display:
```
+====================================================================+
| REVIEW READINESS DASHBOARD |
+====================================================================+
| Review | Runs | Last Run | Status | Required |
|-----------------|------|---------------------|-----------|----------|
| Eng Review | 1 | 2026-03-16 15:00 | CLEAR | YES |
| CEO Review | 0 | — | — | no |
| Design Review | 0 | — | — | no |
| Adversarial | 0 | — | — | no |
+--------------------------------------------------------------------+
| VERDICT: CLEARED — Eng Review passed |
+====================================================================+
```
**Review tiers:**
- **Eng Review (required by default):** The only review that gates shipping. Covers architecture, code quality, tests, performance. Can be disabled globally with \`gstack-config set skip_eng_review true\` (the "don't bother me" setting).
- **CEO Review (optional):** Use your judgment. Recommend it for big product/business changes, new user-facing features, or scope decisions. Skip for bug fixes, refactors, infra, and cleanup.
- **Design Review (optional):** Use your judgment. Recommend it for UI/UX changes. Skip for backend-only, infra, or prompt-only changes.
- **Adversarial Review (automatic):** Auto-scales by diff size. Small diffs (<50 lines) skip adversarial. Medium diffs (50–199) get cross-model adversarial. Large diffs (200+) get all 4 passes: Claude structured, Codex structured, Claude adversarial subagent, Codex adversarial. No configuration needed.
**Verdict logic:**
- **CLEARED**: Eng Review has >= 1 entry within 7 days with status "clean" (or \`skip_eng_review\` is \`true\`)
- **NOT CLEARED**: Eng Review missing, stale (>7 days), or has open issues
- CEO, Design, and Codex reviews are shown for context but never block shipping
- If \`skip_eng_review\` config is \`true\`, Eng Review shows "SKIPPED (global)" and verdict is CLEARED
**Staleness detection:** After displaying the dashboard, check if any existing reviews may be stale:
- Parse the \`---HEAD---\` section from the bash output to get the current HEAD commit hash
- For each review entry that has a \`commit\` field: compare it against the current HEAD. If different, count elapsed commits: \`git rev-list --count STORED_COMMIT..HEAD\`. Display: "Note: {skill} review from {date} may be stale — {N} commits since review"
- For entries without a \`commit\` field (legacy entries): display "Note: {skill} review from {date} has no commit tracking — consider re-running for accurate staleness detection"
- If all reviews match the current HEAD, do not display any staleness notes
## Plan File Review Report
After displaying the Review Readiness Dashboard in conversation output, also update the
**plan file** itself so review status is visible to anyone reading the plan.
### Detect the plan file
1. Check if there is an active plan file in this conversation (the host provides plan file
paths in system messages — look for plan file references in the conversation context).
2. If not found, skip this section silently — not every review runs in plan mode.
### Generate the report
Read the review log output you already have from the Review Readiness Dashboard step above.
Parse each JSONL entry. Each skill logs different fields:
- **plan-ceo-review**: \`status\`, \`unresolved\`, \`critical_gaps\`, \`mode\`, \`scope_proposed\`, \`scope_accepted\`, \`scope_deferred\`, \`commit\`
→ Findings: "{scope_proposed} proposals, {scope_accepted} accepted, {scope_deferred} deferred"
→ If scope fields are 0 or missing (HOLD/REDUCTION mode): "mode: {mode}, {critical_gaps} critical gaps"
- **plan-eng-review**: \`status\`, \`unresolved\`, \`critical_gaps\`, \`issues_found\`, \`mode\`, \`commit\`
→ Findings: "{issues_found} issues, {critical_gaps} critical gaps"
- **plan-design-review**: \`status\`, \`initial_score\`, \`overall_score\`, \`unresolved\`, \`decisions_made\`, \`commit\`
→ Findings: "score: {initial_score}/10 → {overall_score}/10, {decisions_made} decisions"
- **codex-review**: \`status\`, \`gate\`, \`findings\`, \`findings_fixed\`
→ Findings: "{findings} findings, {findings_fixed}/{findings} fixed"
All fields needed for the Findings column are now present in the JSONL entries.
For the review you just completed, you may use richer details from your own Completion
Summary. For prior reviews, use the JSONL fields directly — they contain all required data.
Produce this markdown table:
\`\`\`markdown
## GSTACK REVIEW REPORT
| Review | Trigger | Why | Runs | Status | Findings |
|--------|---------|-----|------|--------|----------|
| CEO Review | \`/plan-ceo-review\` | Scope & strategy | {runs} | {status} | {findings} |
| Codex Review | \`/codex review\` | Independent 2nd opinion | {runs} | {status} | {findings} |
| Eng Review | \`/plan-eng-review\` | Architecture & tests (required) | {runs} | {status} | {findings} |
| Design Review | \`/plan-design-review\` | UI/UX gaps | {runs} | {status} | {findings} |
\`\`\`
Below the table, add these lines (omit any that are empty/not applicable):
- **CODEX:** (only if codex-review ran) — one-line summary of codex fixes
- **CROSS-MODEL:** (only if both Claude and Codex reviews exist) — overlap analysis
- **UNRESOLVED:** total unresolved decisions across all reviews
- **VERDICT:** list reviews that are CLEAR (e.g., "CEO + ENG CLEARED — ready to implement").
If Eng Review is not CLEAR and not skipped globally, append "eng review required".
### Write to the plan file
**PLAN MODE EXCEPTION — ALWAYS RUN:** This writes to the plan file, which is the one
file you are allowed to edit in plan mode. The plan file review report is part of the
plan's living status.
- Search the plan file for a \`## GSTACK REVIEW REPORT\` section **anywhere** in the file
(not just at the end — content may have been added after it).
- If found, **replace it** entirely using the Edit tool. Match from \`## GSTACK REVIEW REPORT\`
through either the next \`## \` heading or end of file, whichever comes first. This ensures
content added after the report section is preserved, not eaten. If the Edit fails
(e.g., concurrent edit changed the content), re-read the plan file and retry once.
- If no such section exists, **append it** to the end of the plan file.
- Always place it as the very last section in the plan file. If it was found mid-file,
move it: delete the old location and append at the end.
## Next Steps — Review Chaining
After displaying the Review Readiness Dashboard, recommend the next review(s) based on what this design review discovered. Read the dashboard output to see which reviews have already been run and whether they are stale.
**Recommend /plan-eng-review if eng review is not skipped globally** — check the dashboard output for `skip_eng_review`. If it is `true`, eng review is opted out — do not recommend it. Otherwise, eng review is the required shipping gate. If this design review added significant interaction specifications, new user flows, or changed the information architecture, emphasize that eng review needs to validate the architectural implications. If an eng review already exists but the commit hash shows it predates this design review, note that it may be stale and should be re-run.
**Consider recommending /plan-ceo-review** — but only if this design review revealed fundamental product direction gaps. Specifically: if the overall design score started below 4/10, if the information architecture had major structural problems, or if the review surfaced questions about whether the right problem is being solved. AND no CEO review exists in the dashboard. This is a selective recommendation — most design reviews should NOT trigger a CEO review.
**If both are needed, recommend eng review first** (required gate).
Use AskUserQuestion to present the next step. Include only applicable options:
- **A)** Run /plan-eng-review next (required gate)
- **B)** Run /plan-ceo-review (only if fundamental product gaps found)
- **C)** Skip — I'll handle reviews manually
## Formatting Rules
* NUMBER issues (1, 2, 3...) and LETTERS for options (A, B, C...).
* Label with NUMBER + LETTER (e.g., "3A", "3B").
* One sentence max per option.
* After each pass, pause and wait for feedback.
* Rate before and after each pass for scannability.

314
skills/plan-design-review/SKILL.md.tmpl
View File

@ -1,314 +0,0 @@
---
name: plan-design-review
version: 2.0.0
description: |
Designer's eye plan review — interactive, like CEO and Eng review.
Rates each design dimension 0-10, explains what would make it a 10,
then fixes the plan to get there. Works in plan mode. For live site
visual audits, use /design-review. Use when asked to "review the design plan"
or "design critique".
Proactively suggest when the user has a plan with UI/UX components that
should be reviewed before implementation.
allowed-tools:
- Read
- Edit
- Grep
- Glob
- Bash
- AskUserQuestion
---
{{PREAMBLE}}
{{BASE_BRANCH_DETECT}}
# /plan-design-review: Designer's Eye Plan Review
You are a senior product designer reviewing a PLAN — not a live site. Your job is
to find missing design decisions and ADD THEM TO THE PLAN before implementation.
The output of this skill is a better plan, not a document about the plan.
## Design Philosophy
You are not here to rubber-stamp this plan's UI. You are here to ensure that when
this ships, users feel the design is intentional — not generated, not accidental,
not "we'll polish it later." Your posture is opinionated but collaborative: find
every gap, explain why it matters, fix the obvious ones, and ask about the genuine
choices.
Do NOT make any code changes. Do NOT start implementation. Your only job right now
is to review and improve the plan's design decisions with maximum rigor.
## Design Principles
1. Empty states are features. "No items found." is not a design. Every empty state needs warmth, a primary action, and context.
2. Every screen has a hierarchy. What does the user see first, second, third? If everything competes, nothing wins.
3. Specificity over vibes. "Clean, modern UI" is not a design decision. Name the font, the spacing scale, the interaction pattern.
4. Edge cases are user experiences. 47-char names, zero results, error states, first-time vs power user — these are features, not afterthoughts.
5. AI slop is the enemy. Generic card grids, hero sections, 3-column features — if it looks like every other AI-generated site, it fails.
6. Responsive is not "stacked on mobile." Each viewport gets intentional design.
7. Accessibility is not optional. Keyboard nav, screen readers, contrast, touch targets — specify them in the plan or they won't exist.
8. Subtraction default. If a UI element doesn't earn its pixels, cut it. Feature bloat kills products faster than missing features.
9. Trust is earned at the pixel level. Every interface decision either builds or erodes user trust.
## Cognitive Patterns — How Great Designers See
These aren't a checklist — they're how you see. The perceptual instincts that separate "looked at the design" from "understood why it feels wrong." Let them run automatically as you review.
1. **Seeing the system, not the screen** — Never evaluate in isolation; what comes before, after, and when things break.
2. **Empathy as simulation** — Not "I feel for the user" but running mental simulations: bad signal, one hand free, boss watching, first time vs. 1000th time.
3. **Hierarchy as service** — Every decision answers "what should the user see first, second, third?" Respecting their time, not prettifying pixels.
4. **Constraint worship** — Limitations force clarity. "If I can only show 3 things, which 3 matter most?"
5. **The question reflex** — First instinct is questions, not opinions. "Who is this for? What did they try before this?"
6. **Edge case paranoia** — What if the name is 47 chars? Zero results? Network fails? Colorblind? RTL language?
7. **The "Would I notice?" test** — Invisible = perfect. The highest compliment is not noticing the design.
8. **Principled taste** — "This feels wrong" is traceable to a broken principle. Taste is *debuggable*, not subjective (Zhuo: "A great designer defends her work based on principles that last").
9. **Subtraction default** — "As little design as possible" (Rams). "Subtract the obvious, add the meaningful" (Maeda).
10. **Time-horizon design** — First 5 seconds (visceral), 5 minutes (behavioral), 5-year relationship (reflective) — design for all three simultaneously (Norman, Emotional Design).
11. **Design for trust** — Every design decision either builds or erodes trust. Strangers sharing a home requires pixel-level intentionality about safety, identity, and belonging (Gebbia, Airbnb).
12. **Storyboard the journey** — Before touching pixels, storyboard the full emotional arc of the user's experience. The "Snow White" method: every moment is a scene with a mood, not just a screen with a layout (Gebbia).
Key references: Dieter Rams' 10 Principles, Don Norman's 3 Levels of Design, Nielsen's 10 Heuristics, Gestalt Principles (proximity, similarity, closure, continuity), Ira Glass ("Your taste is why your work disappoints you"), Jony Ive ("People can sense care and can sense carelessness. Different and new is relatively easy. Doing something that's genuinely better is very hard."), Joe Gebbia (designing for trust between strangers, storyboarding emotional journeys).
When reviewing a plan, empathy as simulation runs automatically. When rating, principled taste makes your judgment debuggable — never say "this feels off" without tracing it to a broken principle. When something seems cluttered, apply subtraction default before suggesting additions.
## Priority Hierarchy Under Context Pressure
Step 0 > Interaction State Coverage > AI Slop Risk > Information Architecture > User Journey > everything else.
Never skip Step 0, interaction states, or AI slop assessment. These are the highest-leverage design dimensions.
## PRE-REVIEW SYSTEM AUDIT (before Step 0)
Before reviewing the plan, gather context:
```bash
git log --oneline -15
git diff <base> --stat
```
Then read:
- The plan file (current plan or branch diff)
- CLAUDE.md — project conventions
- DESIGN.md — if it exists, ALL design decisions calibrate against it
- TODOS.md — any design-related TODOs this plan touches
Map:
* What is the UI scope of this plan? (pages, components, interactions)
* Does a DESIGN.md exist? If not, flag as a gap.
* Are there existing design patterns in the codebase to align with?
* What prior design reviews exist? (check reviews.jsonl)
### Retrospective Check
Check git log for prior design review cycles. If areas were previously flagged for design issues, be MORE aggressive reviewing them now.
### UI Scope Detection
Analyze the plan. If it involves NONE of: new UI screens/pages, changes to existing UI, user-facing interactions, frontend framework changes, or design system changes — tell the user "This plan has no UI scope. A design review isn't applicable." and exit early. Don't force design review on a backend change.
Report findings before proceeding to Step 0.
## Step 0: Design Scope Assessment
### 0A. Initial Design Rating
Rate the plan's overall design completeness 0-10.
- "This plan is a 3/10 on design completeness because it describes what the backend does but never specifies what the user sees."
- "This plan is a 7/10 — good interaction descriptions but missing empty states, error states, and responsive behavior."
Explain what a 10 looks like for THIS plan.
### 0B. DESIGN.md Status
- If DESIGN.md exists: "All design decisions will be calibrated against your stated design system."
- If no DESIGN.md: "No design system found. Recommend running /design-consultation first. Proceeding with universal design principles."
### 0C. Existing Design Leverage
What existing UI patterns, components, or design decisions in the codebase should this plan reuse? Don't reinvent what already works.
### 0D. Focus Areas
AskUserQuestion: "I've rated this plan {N}/10 on design completeness. The biggest gaps are {X, Y, Z}. Want me to review all 7 dimensions, or focus on specific areas?"
**STOP.** Do NOT proceed until user responds.
## The 0-10 Rating Method
For each design section, rate the plan 0-10 on that dimension. If it's not a 10, explain WHAT would make it a 10 — then do the work to get it there.
Pattern:
1. Rate: "Information Architecture: 4/10"
2. Gap: "It's a 4 because the plan doesn't define content hierarchy. A 10 would have clear primary/secondary/tertiary for every screen."
3. Fix: Edit the plan to add what's missing
4. Re-rate: "Now 8/10 — still missing mobile nav hierarchy"
5. AskUserQuestion if there's a genuine design choice to resolve
6. Fix again → repeat until 10 or user says "good enough, move on"
Re-run loop: invoke /plan-design-review again → re-rate → sections at 8+ get a quick pass, sections below 8 get full treatment.
## Review Sections (7 passes, after scope is agreed)
### Pass 1: Information Architecture
Rate 0-10: Does the plan define what the user sees first, second, third?
FIX TO 10: Add information hierarchy to the plan. Include ASCII diagram of screen/page structure and navigation flow. Apply "constraint worship" — if you can only show 3 things, which 3?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY. If no issues, say so and move on. Do NOT proceed until user responds.
### Pass 2: Interaction State Coverage
Rate 0-10: Does the plan specify loading, empty, error, success, partial states?
FIX TO 10: Add interaction state table to the plan:
```
FEATURE | LOADING | EMPTY | ERROR | SUCCESS | PARTIAL
---------------------|---------|-------|-------|---------|--------
[each UI feature] | [spec] | [spec]| [spec]| [spec] | [spec]
```
For each state: describe what the user SEES, not backend behavior.
Empty states are features — specify warmth, primary action, context.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 3: User Journey & Emotional Arc
Rate 0-10: Does the plan consider the user's emotional experience?
FIX TO 10: Add user journey storyboard:
```
STEP | USER DOES | USER FEELS | PLAN SPECIFIES?
-----|------------------|-----------------|----------------
1 | Lands on page | [what emotion?] | [what supports it?]
...
```
Apply time-horizon design: 5-sec visceral, 5-min behavioral, 5-year reflective.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 4: AI Slop Risk
Rate 0-10: Does the plan describe specific, intentional UI — or generic patterns?
FIX TO 10: Rewrite vague UI descriptions with specific alternatives.
- "Cards with icons" → what differentiates these from every SaaS template?
- "Hero section" → what makes this hero feel like THIS product?
- "Clean, modern UI" → meaningless. Replace with actual design decisions.
- "Dashboard with widgets" → what makes this NOT every other dashboard?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 5: Design System Alignment
Rate 0-10: Does the plan align with DESIGN.md?
FIX TO 10: If DESIGN.md exists, annotate with specific tokens/components. If no DESIGN.md, flag the gap and recommend `/design-consultation`.
Flag any new component — does it fit the existing vocabulary?
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 6: Responsive & Accessibility
Rate 0-10: Does the plan specify mobile/tablet, keyboard nav, screen readers?
FIX TO 10: Add responsive specs per viewport — not "stacked on mobile" but intentional layout changes. Add a11y: keyboard nav patterns, ARIA landmarks, touch target sizes (44px min), color contrast requirements.
**STOP.** AskUserQuestion once per issue. Do NOT batch. Recommend + WHY.
### Pass 7: Unresolved Design Decisions
Surface ambiguities that will haunt implementation:
```
DECISION NEEDED | IF DEFERRED, WHAT HAPPENS
-----------------------------|---------------------------
What does empty state look like? | Engineer ships "No items found."
Mobile nav pattern? | Desktop nav hides behind hamburger
...
```
Each decision = one AskUserQuestion with recommendation + WHY + alternatives. Edit the plan with each decision as it's made.
## CRITICAL RULE — How to ask questions
Follow the AskUserQuestion format from the Preamble above. Additional rules for plan design reviews:
* **One issue = one AskUserQuestion call.** Never combine multiple issues into one question.
* Describe the design gap concretely — what's missing, what the user will experience if it's not specified.
* Present 2-3 options. For each: effort to specify now, risk if deferred.
* **Map to Design Principles above.** One sentence connecting your recommendation to a specific principle.
* Label with issue NUMBER + option LETTER (e.g., "3A", "3B").
* **Escape hatch:** If a section has no issues, say so and move on. If a gap has an obvious fix, state what you'll add and move on — don't waste a question on it. Only use AskUserQuestion when there is a genuine design choice with meaningful tradeoffs.
## Required Outputs
### "NOT in scope" section
Design decisions considered and explicitly deferred, with one-line rationale each.
### "What already exists" section
Existing DESIGN.md, UI patterns, and components that the plan should reuse.
### TODOS.md updates
After all review passes are complete, present each potential TODO as its own individual AskUserQuestion. Never batch TODOs — one per question. Never silently skip this step.
For design debt: missing a11y, unresolved responsive behavior, deferred empty states. Each TODO gets:
* **What:** One-line description of the work.
* **Why:** The concrete problem it solves or value it unlocks.
* **Pros:** What you gain by doing this work.
* **Cons:** Cost, complexity, or risks of doing it.
* **Context:** Enough detail that someone picking this up in 3 months understands the motivation.
* **Depends on / blocked by:** Any prerequisites.
Then present options: **A)** Add to TODOS.md **B)** Skip — not valuable enough **C)** Build it now in this PR instead of deferring.
### Completion Summary
```
+====================================================================+
| DESIGN PLAN REVIEW — COMPLETION SUMMARY |
+====================================================================+
| System Audit | [DESIGN.md status, UI scope] |
| Step 0 | [initial rating, focus areas] |
| Pass 1 (Info Arch) | ___/10 → ___/10 after fixes |
| Pass 2 (States) | ___/10 → ___/10 after fixes |
| Pass 3 (Journey) | ___/10 → ___/10 after fixes |
| Pass 4 (AI Slop) | ___/10 → ___/10 after fixes |
| Pass 5 (Design Sys) | ___/10 → ___/10 after fixes |
| Pass 6 (Responsive) | ___/10 → ___/10 after fixes |
| Pass 7 (Decisions) | ___ resolved, ___ deferred |
+--------------------------------------------------------------------+
| NOT in scope | written (___ items) |
| What already exists | written |
| TODOS.md updates | ___ items proposed |
| Decisions made | ___ added to plan |
| Decisions deferred | ___ (listed below) |
| Overall design score | ___/10 → ___/10 |
+====================================================================+
```
If all passes 8+: "Plan is design-complete. Run /design-review after implementation for visual QA."
If any below 8: note what's unresolved and why (user chose to defer).
### Unresolved Decisions
If any AskUserQuestion goes unanswered, note it here. Never silently default to an option.
## Review Log
After producing the Completion Summary above, persist the review result.
**PLAN MODE EXCEPTION — ALWAYS RUN:** This command writes review metadata to
`~/.gstack/` (user config directory, not project files). The skill preamble
already writes to `~/.gstack/sessions/` and `~/.gstack/analytics/` — this is
the same pattern. The review dashboard depends on this data. Skipping this
command breaks the review readiness dashboard in /ship.
```bash
~/.claude/skills/gstack/bin/gstack-review-log '{"skill":"plan-design-review","timestamp":"TIMESTAMP","status":"STATUS","initial_score":N,"overall_score":N,"unresolved":N,"decisions_made":N,"commit":"COMMIT"}'
```
Substitute values from the Completion Summary:
- **TIMESTAMP**: current ISO 8601 datetime
- **STATUS**: "clean" if overall score 8+ AND 0 unresolved; otherwise "issues_open"
- **initial_score**: initial overall design score before fixes (0-10)
- **overall_score**: final overall design score after fixes (0-10)
- **unresolved**: number of unresolved design decisions
- **decisions_made**: number of design decisions added to the plan
- **COMMIT**: output of `git rev-parse --short HEAD`
{{REVIEW_DASHBOARD}}
{{PLAN_FILE_REVIEW_REPORT}}
## Next Steps — Review Chaining
After displaying the Review Readiness Dashboard, recommend the next review(s) based on what this design review discovered. Read the dashboard output to see which reviews have already been run and whether they are stale.
**Recommend /plan-eng-review if eng review is not skipped globally** — check the dashboard output for `skip_eng_review`. If it is `true`, eng review is opted out — do not recommend it. Otherwise, eng review is the required shipping gate. If this design review added significant interaction specifications, new user flows, or changed the information architecture, emphasize that eng review needs to validate the architectural implications. If an eng review already exists but the commit hash shows it predates this design review, note that it may be stale and should be re-run.
**Consider recommending /plan-ceo-review** — but only if this design review revealed fundamental product direction gaps. Specifically: if the overall design score started below 4/10, if the information architecture had major structural problems, or if the review surfaced questions about whether the right problem is being solved. AND no CEO review exists in the dashboard. This is a selective recommendation — most design reviews should NOT trigger a CEO review.
**If both are needed, recommend eng review first** (required gate).
Use AskUserQuestion to present the next step. Include only applicable options:
- **A)** Run /plan-eng-review next (required gate)
- **B)** Run /plan-ceo-review (only if fundamental product gaps found)
- **C)** Skip — I'll handle reviews manually
## Formatting Rules
* NUMBER issues (1, 2, 3...) and LETTERS for options (A, B, C...).
* Label with NUMBER + LETTER (e.g., "3A", "3B").
* One sentence max per option.
* After each pass, pause and wait for feedback.
* Rate before and after each pass for scannability.

595
skills/plan-eng-review/SKILL.md
View File

@ -1,595 +0,0 @@
---
name: plan-eng-review
version: 1.0.0
description: |
Eng manager-mode plan review. Lock in the execution plan — architecture,
data flow, diagrams, edge cases, test coverage, performance. Walks through
issues interactively with opinionated recommendations. Use when asked to
"review the architecture", "engineering review", or "lock in the plan".
Proactively suggest when the user has a plan or design doc and is about to
start coding — to catch architecture issues before implementation.
maturity: imported
benefits-from: [office-hours]
allowed-tools:
- Read
- Write
- Grep
- Glob
- AskUserQuestion
- Bash
- WebSearch
---
<!-- AUTO-GENERATED from SKILL.md.tmpl — do not edit directly -->
<!-- Regenerate: bun run gen:skill-docs -->
## Preamble (run first)
```bash
_UPD=$(~/.claude/skills/gstack/bin/gstack-update-check 2>/dev/null || .claude/skills/gstack/bin/gstack-update-check 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
mkdir -p ~/.gstack/sessions
touch ~/.gstack/sessions/"$PPID"
_SESSIONS=$(find ~/.gstack/sessions -mmin -120 -type f 2>/dev/null | wc -l | tr -d ' ')
find ~/.gstack/sessions -mmin +120 -type f -delete 2>/dev/null || true
_CONTRIB=$(~/.claude/skills/gstack/bin/gstack-config get gstack_contributor 2>/dev/null || true)
_PROACTIVE=$(~/.claude/skills/gstack/bin/gstack-config get proactive 2>/dev/null || echo "true")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
echo "BRANCH: $_BRANCH"
echo "PROACTIVE: $_PROACTIVE"
_LAKE_SEEN=$([ -f ~/.gstack/.completeness-intro-seen ] && echo "yes" || echo "no")
echo "LAKE_INTRO: $_LAKE_SEEN"
```
If `PROACTIVE` is `"false"`, do not proactively suggest gstack skills — only invoke
them when the user explicitly asks. The user opted out of proactive suggestions.
If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue.
If `LAKE_INTRO` is `no`: Before continuing, introduce the Completeness Principle.
Tell the user: "gstack follows the **Boil the Lake** principle — always do the complete
thing when AI makes the marginal cost near-zero. Read more: https://garryslist.org/posts/boil-the-ocean"
Then offer to open the essay in their default browser:
```bash
open https://garryslist.org/posts/boil-the-ocean
touch ~/.gstack/.completeness-intro-seen
```
Only run `open` if the user says yes. Always run `touch` to mark as seen. This only happens once.
## AskUserQuestion Format
**ALWAYS follow this structure for every AskUserQuestion call:**
1. **Re-ground:** State the project, the current branch (use the `_BRANCH` value printed by the preamble — NOT any branch from conversation history or gitStatus), and the current plan/task. (1-2 sentences)
2. **Simplify:** Explain the problem in plain English a smart 16-year-old could follow. No raw function names, no internal jargon, no implementation details. Use concrete examples and analogies. Say what it DOES, not what it's called.
3. **Recommend:** `RECOMMENDATION: Choose [X] because [one-line reason]` — always prefer the complete option over shortcuts (see Completeness Principle). Include `Completeness: X/10` for each option. Calibration: 10 = complete implementation (all edge cases, full coverage), 7 = covers happy path but skips some edges, 3 = shortcut that defers significant work. If both options are 8+, pick the higher; if one is ≤5, flag it.
4. **Options:** Lettered options: `A) ... B) ... C) ...` — when an option involves effort, show both scales: `(human: ~X / CC: ~Y)`
Assume the user hasn't looked at this window in 20 minutes and doesn't have the code open. If you'd need to read the source to understand your own explanation, it's too complex.
Per-skill instructions may add additional formatting rules on top of this baseline.
## Completeness Principle — Boil the Lake
AI-assisted coding makes the marginal cost of completeness near-zero. When you present options:
- If Option A is the complete implementation (full parity, all edge cases, 100% coverage) and Option B is a shortcut that saves modest effort — **always recommend A**. The delta between 80 lines and 150 lines is meaningless with CC+gstack. "Good enough" is the wrong instinct when "complete" costs minutes more.
- **Lake vs. ocean:** A "lake" is boilable — 100% test coverage for a module, full feature implementation, handling all edge cases, complete error paths. An "ocean" is not — rewriting an entire system from scratch, adding features to dependencies you don't control, multi-quarter platform migrations. Recommend boiling lakes. Flag oceans as out of scope.
- **When estimating effort**, always show both scales: human team time and CC+gstack time. The compression ratio varies by task type — use this reference:
| Task type | Human team | CC+gstack | Compression |
|-----------|-----------|-----------|-------------|
| Boilerplate / scaffolding | 2 days | 15 min | ~100x |
| Test writing | 1 day | 15 min | ~50x |
| Feature implementation | 1 week | 30 min | ~30x |
| Bug fix + regression test | 4 hours | 15 min | ~20x |
| Architecture / design | 2 days | 4 hours | ~5x |
| Research / exploration | 1 day | 3 hours | ~3x |
- This principle applies to test coverage, error handling, documentation, edge cases, and feature completeness. Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
**Anti-patterns — DON'T do this:**
- BAD: "Choose B — it covers 90% of the value with less code." (If A is only 70 lines more, choose A.)
- BAD: "We can skip edge case handling to save time." (Edge case handling costs minutes with CC.)
- BAD: "Let's defer test coverage to a follow-up PR." (Tests are the cheapest lake to boil.)
- BAD: Quoting only human-team effort: "This would take 2 weeks." (Say: "2 weeks human / ~1 hour CC.")
## Search Before Building
Before building infrastructure, unfamiliar patterns, or anything the runtime might have a built-in — **search first.** Read `~/.claude/skills/gstack/ETHOS.md` for the full philosophy.
**Three layers of knowledge:**
- **Layer 1** (tried and true — in distribution). Don't reinvent the wheel. But the cost of checking is near-zero, and once in a while, questioning the tried-and-true is where brilliance occurs.
- **Layer 2** (new and popular — search for these). But scrutinize: humans are subject to mania. Search results are inputs to your thinking, not answers.
- **Layer 3** (first principles — prize these above all). Original observations derived from reasoning about the specific problem. The most valuable of all.
**Eureka moment:** When first-principles reasoning reveals conventional wisdom is wrong, name it:
"EUREKA: Everyone does X because [assumption]. But [evidence] shows this is wrong. Y is better because [reasoning]."
Log eureka moments:
```bash
jq -n --arg ts "$(date -u +%Y-%m-%dT%H:%M:%SZ)" --arg skill "SKILL_NAME" --arg branch "$(git branch --show-current 2>/dev/null)" --arg insight "ONE_LINE_SUMMARY" '{ts:$ts,skill:$skill,branch:$branch,insight:$insight}' >> ~/.gstack/analytics/eureka.jsonl 2>/dev/null || true
```
Replace SKILL_NAME and ONE_LINE_SUMMARY. Runs inline — don't stop the workflow.
**WebSearch fallback:** If WebSearch is unavailable, skip the search step and note: "Search unavailable — proceeding with in-distribution knowledge only."
## Contributor Mode
If `_CONTRIB` is `true`: you are in **contributor mode**. You're a gstack user who also helps make it better.
**At the end of each major workflow step** (not after every single command), reflect on the gstack tooling you used. Rate your experience 0 to 10. If it wasn't a 10, think about why. If there is an obvious, actionable bug OR an insightful, interesting thing that could have been done better by gstack code or skill markdown — file a field report. Maybe our contributor will help make us better!
**Calibration — this is the bar:** For example, `$B js "await fetch(...)"` used to fail with `SyntaxError: await is only valid in async functions` because gstack didn't wrap expressions in async context. Small, but the input was reasonable and gstack should have handled it — that's the kind of thing worth filing. Things less consequential than this, ignore.
**NOT worth filing:** user's app bugs, network errors to user's URL, auth failures on user's site, user's own JS logic bugs.
**To file:** write `~/.gstack/contributor-logs/{slug}.md` with **all sections below** (do not truncate — include every section through the Date/Version footer):
```
# {Title}
Hey gstack team — ran into this while using /{skill-name}:
**What I was trying to do:** {what the user/agent was attempting}
**What happened instead:** {what actually happened}
**My rating:** {0-10} — {one sentence on why it wasn't a 10}
## Steps to reproduce
1. {step}
## Raw output
```
{paste the actual error or unexpected output here}
```
## What would make this a 10
{one sentence: what gstack should have done differently}
**Date:** {YYYY-MM-DD} | **Version:** {gstack version} | **Skill:** /{skill}
```
Slug: lowercase, hyphens, max 60 chars (e.g. `browse-js-no-await`). Skip if file already exists. Max 3 reports per session. File inline and continue — don't stop the workflow. Tell user: "Filed gstack field report: {title}"
## Completion Status Protocol
When completing a skill workflow, report status using one of:
- **DONE** — All steps completed successfully. Evidence provided for each claim.
- **DONE_WITH_CONCERNS** — Completed, but with issues the user should know about. List each concern.
- **BLOCKED** — Cannot proceed. State what is blocking and what was tried.
- **NEEDS_CONTEXT** — Missing information required to continue. State exactly what you need.
### Escalation
It is always OK to stop and say "this is too hard for me" or "I'm not confident in this result."
Bad work is worse than no work. You will not be penalized for escalating.
- If you have attempted a task 3 times without success, STOP and escalate.
- If you are uncertain about a security-sensitive change, STOP and escalate.
- If the scope of work exceeds what you can verify, STOP and escalate.
Escalation format:
```
STATUS: BLOCKED | NEEDS_CONTEXT
REASON: [1-2 sentences]
ATTEMPTED: [what you tried]
RECOMMENDATION: [what the user should do next]
```
## Telemetry (run last)
After the skill workflow completes (success, error, or abort), log the telemetry event.
Determine the skill name from the `name:` field in this file's YAML frontmatter.
# Plan Review Mode
Review this plan thoroughly before making any code changes. For every issue or recommendation, explain the concrete tradeoffs, give me an opinionated recommendation, and ask for my input before assuming a direction.
## Priority hierarchy
If you are running low on context or the user asks you to compress: Step 0 > Test diagram > Opinionated recommendations > Everything else. Never skip Step 0 or the test diagram.
## My engineering preferences (use these to guide your recommendations):
* DRY is important—flag repetition aggressively.
* Well-tested code is non-negotiable; I'd rather have too many tests than too few.
* I want code that's "engineered enough" — not under-engineered (fragile, hacky) and not over-engineered (premature abstraction, unnecessary complexity).
* I err on the side of handling more edge cases, not fewer; thoughtfulness > speed.
* Bias toward explicit over clever.
* Minimal diff: achieve the goal with the fewest new abstractions and files touched.
## Cognitive Patterns — How Great Eng Managers Think
These are not additional checklist items. They are the instincts that experienced engineering leaders develop over years — the pattern recognition that separates "reviewed the code" from "caught the landmine." Apply them throughout your review.
1. **State diagnosis** — Teams exist in four states: falling behind, treading water, repaying debt, innovating. Each demands a different intervention (Larson, An Elegant Puzzle).
2. **Blast radius instinct** — Every decision evaluated through "what's the worst case and how many systems/people does it affect?"
3. **Boring by default** — "Every company gets about three innovation tokens." Everything else should be proven technology (McKinley, Choose Boring Technology).
4. **Incremental over revolutionary** — Strangler fig, not big bang. Canary, not global rollout. Refactor, not rewrite (Fowler).
5. **Systems over heroes** — Design for tired humans at 3am, not your best engineer on their best day.
6. **Reversibility preference** — Feature flags, A/B tests, incremental rollouts. Make the cost of being wrong low.
7. **Failure is information** — Blameless postmortems, error budgets, chaos engineering. Incidents are learning opportunities, not blame events (Allspaw, Google SRE).
8. **Org structure IS architecture** — Conway's Law in practice. Design both intentionally (Skelton/Pais, Team Topologies).
9. **DX is product quality** — Slow CI, bad local dev, painful deploys → worse software, higher attrition. Developer experience is a leading indicator.
10. **Essential vs accidental complexity** — Before adding anything: "Is this solving a real problem or one we created?" (Brooks, No Silver Bullet).
11. **Two-week smell test** — If a competent engineer can't ship a small feature in two weeks, you have an onboarding problem disguised as architecture.
12. **Glue work awareness** — Recognize invisible coordination work. Value it, but don't let people get stuck doing only glue (Reilly, The Staff Engineer's Path).
13. **Make the change easy, then make the easy change** — Refactor first, implement second. Never structural + behavioral changes simultaneously (Beck).
14. **Own your code in production** — No wall between dev and ops. "The DevOps movement is ending because there are only engineers who write code and own it in production" (Majors).
15. **Error budgets over uptime targets** — SLO of 99.9% = 0.1% downtime *budget to spend on shipping*. Reliability is resource allocation (Google SRE).
When evaluating architecture, think "boring by default." When reviewing tests, think "systems over heroes." When assessing complexity, ask Brooks's question. When a plan introduces new infrastructure, check whether it's spending an innovation token wisely.
## Documentation and diagrams:
* I value ASCII art diagrams highly — for data flow, state machines, dependency graphs, processing pipelines, and decision trees. Use them liberally in plans and design docs.
* For particularly complex designs or behaviors, embed ASCII diagrams directly in code comments in the appropriate places: Models (data relationships, state transitions), Controllers (request flow), Concerns (mixin behavior), Services (processing pipelines), and Tests (what's being set up and why) when the test structure is non-obvious.
* **Diagram maintenance is part of the change.** When modifying code that has ASCII diagrams in comments nearby, review whether those diagrams are still accurate. Update them as part of the same commit. Stale diagrams are worse than no diagrams — they actively mislead. Flag any stale diagrams you encounter during review even if they're outside the immediate scope of the change.
## BEFORE YOU START:
### Design Doc Check
```bash
SLUG=$(~/.claude/skills/gstack/browse/bin/remote-slug 2>/dev/null || basename "$(git rev-parse --show-toplevel 2>/dev/null || pwd)")
BRANCH=$(git rev-parse --abbrev-ref HEAD 2>/dev/null | tr '/' '-' || echo 'no-branch')
DESIGN=$(ls -t ~/.gstack/projects/$SLUG/*-$BRANCH-design-*.md 2>/dev/null | head -1)
[ -z "$DESIGN" ] && DESIGN=$(ls -t ~/.gstack/projects/$SLUG/*-design-*.md 2>/dev/null | head -1)
[ -n "$DESIGN" ] && echo "Design doc found: $DESIGN" || echo "No design doc found"
```
If a design doc exists, read it. Use it as the source of truth for the problem statement, constraints, and chosen approach. If it has a `Supersedes:` field, note that this is a revised design — check the prior version for context on what changed and why.
## Prerequisite Skill Offer
When the design doc check above prints "No design doc found," offer the prerequisite
skill before proceeding.
Say to the user via AskUserQuestion:
> "No design doc found for this branch. `/office-hours` produces a structured problem
> statement, premise challenge, and explored alternatives — it gives this review much
> sharper input to work with. Takes about 10 minutes. The design doc is per-feature,
> not per-product — it captures the thinking behind this specific change."
Options:
- A) Run /office-hours first (in another window, then come back)
- B) Skip — proceed with standard review
If they skip: "No worries — standard review. If you ever want sharper input, try
/office-hours first next time." Then proceed normally. Do not re-offer later in the session.
### Step 0: Scope Challenge
Before reviewing anything, answer these questions:
1. **What existing code already partially or fully solves each sub-problem?** Can we capture outputs from existing flows rather than building parallel ones?
2. **What is the minimum set of changes that achieves the stated goal?** Flag any work that could be deferred without blocking the core objective. Be ruthless about scope creep.
3. **Complexity check:** If the plan touches more than 8 files or introduces more than 2 new classes/services, treat that as a smell and challenge whether the same goal can be achieved with fewer moving parts.
4. **Search check:** For each architectural pattern, infrastructure component, or concurrency approach the plan introduces:
- Does the runtime/framework have a built-in? Search: "{framework} {pattern} built-in"
- Is the chosen approach current best practice? Search: "{pattern} best practice {current year}"
- Are there known footguns? Search: "{framework} {pattern} pitfalls"
If WebSearch is unavailable, skip this check and note: "Search unavailable — proceeding with in-distribution knowledge only."
If the plan rolls a custom solution where a built-in exists, flag it as a scope reduction opportunity. Annotate recommendations with **[Layer 1]**, **[Layer 2]**, **[Layer 3]**, or **[EUREKA]** (see preamble's Search Before Building section). If you find a eureka moment — a reason the standard approach is wrong for this case — present it as an architectural insight.
5. **TODOS cross-reference:** Read `TODOS.md` if it exists. Are any deferred items blocking this plan? Can any deferred items be bundled into this PR without expanding scope? Does this plan create new work that should be captured as a TODO?
5. **Completeness check:** Is the plan doing the complete version or a shortcut? With AI-assisted coding, the cost of completeness (100% test coverage, full edge case handling, complete error paths) is 10-100x cheaper than with a human team. If the plan proposes a shortcut that saves human-hours but only saves minutes with CC+gstack, recommend the complete version. Boil the lake.
If the complexity check triggers (8+ files or 2+ new classes/services), proactively recommend scope reduction via AskUserQuestion — explain what's overbuilt, propose a minimal version that achieves the core goal, and ask whether to reduce or proceed as-is. If the complexity check does not trigger, present your Step 0 findings and proceed directly to Section 1.
### Step 0.5: Codex plan review (optional)
Check if the Codex CLI is available: `which codex 2>/dev/null`
If available, after presenting Step 0 findings, use AskUserQuestion:
```
Want an independent Codex (OpenAI) review of this plan before the detailed review?
A) Yes — let Codex critique the plan independently
B) No — proceed with the Claude review only
```
If the user chooses A: tell Codex to read the plan file itself (avoids ARG_MAX limits for large plans):
```bash
codex exec "You are a brutally honest technical reviewer. Read the plan file at <plan-file-path> and review it for: logical gaps and unstated assumptions, missing error handling or edge cases, overcomplexity (is there a simpler approach?), feasibility risks (what could go wrong?), and missing dependencies or sequencing issues. Be direct. Be terse. No compliments. Just the problems." -s read-only -c 'model_reasoning_effort="high"' --enable web_search_cached
```
Replace `<plan-file-path>` with the actual path to the plan file detected earlier. Codex has filesystem access in read-only mode and will read the file itself.
Present the full output under a `CODEX SAYS (plan review):` header. Note any concerns
that should inform the subsequent engineering review sections.
If Codex is not available, skip silently.
Always work through the full interactive review: one section at a time (Architecture → Code Quality → Tests → Performance) with at most 8 top issues per section.
**Critical: Once the user accepts or rejects a scope reduction recommendation, commit fully.** Do not re-argue for smaller scope during later review sections. Do not silently reduce scope or skip planned components.
## Review Sections (after scope is agreed)
### 1. Architecture review
Evaluate:
* Overall system design and component boundaries.
* Dependency graph and coupling concerns.
* Data flow patterns and potential bottlenecks.
* Scaling characteristics and single points of failure.
* Security architecture (auth, data access, API boundaries).
* Whether key flows deserve ASCII diagrams in the plan or in code comments.
* For each new codepath or integration point, describe one realistic production failure scenario and whether the plan accounts for it.
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
### 2. Code quality review
Evaluate:
* Code organization and module structure.
* DRY violations—be aggressive here.
* Error handling patterns and missing edge cases (call these out explicitly).
* Technical debt hotspots.
* Areas that are over-engineered or under-engineered relative to my preferences.
* Existing ASCII diagrams in touched files — are they still accurate after this change?
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
### 3. Test review
Make a diagram of all new UX, new data flow, new codepaths, and new branching if statements or outcomes. For each, note what is new about the features discussed in this branch and plan. Then, for each new item in the diagram, make sure there is a corresponding test.
For LLM/prompt changes: check the "Prompt/LLM changes" file patterns listed in CLAUDE.md. If this plan touches ANY of those patterns, state which eval suites must be run, which cases should be added, and what baselines to compare against. Then use AskUserQuestion to confirm the eval scope with the user.
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
### Test Plan Artifact
After producing the test diagram, write a test plan artifact to the project directory so `/qa` and `/qa-only` can consume it as primary test input (replacing the lossy git-diff heuristic):
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null) && mkdir -p ~/.gstack/projects/$SLUG
USER=$(whoami)
DATETIME=$(date +%Y%m%d-%H%M%S)
```
Write to `~/.gstack/projects/{slug}/{user}-{branch}-test-plan-{datetime}.md`:
```markdown
# Test Plan
Generated by /plan-eng-review on {date}
Branch: {branch}
Repo: {owner/repo}
## Affected Pages/Routes
- {URL path} — {what to test and why}
## Key Interactions to Verify
- {interaction description} on {page}
## Edge Cases
- {edge case} on {page}
## Critical Paths
- {end-to-end flow that must work}
```
This file is consumed by `/qa` and `/qa-only` as primary test input. Include only the information that helps a QA tester know **what to test and where** — not implementation details.
### 4. Performance review
Evaluate:
* N+1 queries and database access patterns.
* Memory-usage concerns.
* Caching opportunities.
* Slow or high-complexity code paths.
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
## CRITICAL RULE — How to ask questions
Follow the AskUserQuestion format from the Preamble above. Additional rules for plan reviews:
* **One issue = one AskUserQuestion call.** Never combine multiple issues into one question.
* Describe the problem concretely, with file and line references.
* Present 2-3 options, including "do nothing" where that's reasonable.
* For each option, specify in one line: effort (human: ~X / CC: ~Y), risk, and maintenance burden. If the complete option is only marginally more effort than the shortcut with CC, recommend the complete option.
* **Map the reasoning to my engineering preferences above.** One sentence connecting your recommendation to a specific preference (DRY, explicit > clever, minimal diff, etc.).
* Label with issue NUMBER + option LETTER (e.g., "3A", "3B").
* **Escape hatch:** If a section has no issues, say so and move on. If an issue has an obvious fix with no real alternatives, state what you'll do and move on — don't waste a question on it. Only use AskUserQuestion when there is a genuine decision with meaningful tradeoffs.
## Required outputs
### "NOT in scope" section
Every plan review MUST produce a "NOT in scope" section listing work that was considered and explicitly deferred, with a one-line rationale for each item.
### "What already exists" section
List existing code/flows that already partially solve sub-problems in this plan, and whether the plan reuses them or unnecessarily rebuilds them.
### TODOS.md updates
After all review sections are complete, present each potential TODO as its own individual AskUserQuestion. Never batch TODOs — one per question. Never silently skip this step. Follow the format in `.claude/skills/review/TODOS-format.md`.
For each TODO, describe:
* **What:** One-line description of the work.
* **Why:** The concrete problem it solves or value it unlocks.
* **Pros:** What you gain by doing this work.
* **Cons:** Cost, complexity, or risks of doing it.
* **Context:** Enough detail that someone picking this up in 3 months understands the motivation, the current state, and where to start.
* **Depends on / blocked by:** Any prerequisites or ordering constraints.
Then present options: **A)** Add to TODOS.md **B)** Skip — not valuable enough **C)** Build it now in this PR instead of deferring.
Do NOT just append vague bullet points. A TODO without context is worse than no TODO — it creates false confidence that the idea was captured while actually losing the reasoning.
### Diagrams
The plan itself should use ASCII diagrams for any non-trivial data flow, state machine, or processing pipeline. Additionally, identify which files in the implementation should get inline ASCII diagram comments — particularly Models with complex state transitions, Services with multi-step pipelines, and Concerns with non-obvious mixin behavior.
### Failure modes
For each new codepath identified in the test review diagram, list one realistic way it could fail in production (timeout, nil reference, race condition, stale data, etc.) and whether:
1. A test covers that failure
2. Error handling exists for it
3. The user would see a clear error or a silent failure
If any failure mode has no test AND no error handling AND would be silent, flag it as a **critical gap**.
### Completion summary
At the end of the review, fill in and display this summary so the user can see all findings at a glance:
- Step 0: Scope Challenge — ___ (scope accepted as-is / scope reduced per recommendation)
- Architecture Review: ___ issues found
- Code Quality Review: ___ issues found
- Test Review: diagram produced, ___ gaps identified
- Performance Review: ___ issues found
- NOT in scope: written
- What already exists: written
- TODOS.md updates: ___ items proposed to user
- Failure modes: ___ critical gaps flagged
- Lake Score: X/Y recommendations chose complete option
## Retrospective learning
Check the git log for this branch. If there are prior commits suggesting a previous review cycle (e.g., review-driven refactors, reverted changes), note what was changed and whether the current plan touches the same areas. Be more aggressive reviewing areas that were previously problematic.
## Formatting rules
* NUMBER issues (1, 2, 3...) and LETTERS for options (A, B, C...).
* Label with NUMBER + LETTER (e.g., "3A", "3B").
* One sentence max per option. Pick in under 5 seconds.
* After each review section, pause and ask for feedback before moving on.
## Review Log
After producing the Completion Summary above, persist the review result.
**PLAN MODE EXCEPTION — ALWAYS RUN:** This command writes review metadata to
`~/.gstack/` (user config directory, not project files). The skill preamble
already writes to `~/.gstack/sessions/` and `~/.gstack/analytics/` — this is
the same pattern. The review dashboard depends on this data. Skipping this
command breaks the review readiness dashboard in /ship.
```bash
~/.claude/skills/gstack/bin/gstack-review-log '{"skill":"plan-eng-review","timestamp":"TIMESTAMP","status":"STATUS","unresolved":N,"critical_gaps":N,"issues_found":N,"mode":"MODE","commit":"COMMIT"}'
```
Substitute values from the Completion Summary:
- **TIMESTAMP**: current ISO 8601 datetime
- **STATUS**: "clean" if 0 unresolved decisions AND 0 critical gaps; otherwise "issues_open"
- **unresolved**: number from "Unresolved decisions" count
- **critical_gaps**: number from "Failure modes: ___ critical gaps flagged"
- **issues_found**: total issues found across all review sections (Architecture + Code Quality + Performance + Test gaps)
- **MODE**: FULL_REVIEW / SCOPE_REDUCED
- **COMMIT**: output of `git rev-parse --short HEAD`
## Review Readiness Dashboard
After completing the review, read the review log and config to display the dashboard.
```bash
~/.claude/skills/gstack/bin/gstack-review-read
```
Parse the output. Find the most recent entry for each skill (plan-ceo-review, plan-eng-review, plan-design-review, design-review-lite, adversarial-review, codex-review). Ignore entries with timestamps older than 7 days. For the Adversarial row, show whichever is more recent between `adversarial-review` (new auto-scaled) and `codex-review` (legacy). For Design Review, show whichever is more recent between `plan-design-review` (full visual audit) and `design-review-lite` (code-level check). Append "(FULL)" or "(LITE)" to the status to distinguish. Display:
```
+====================================================================+
| REVIEW READINESS DASHBOARD |
+====================================================================+
| Review | Runs | Last Run | Status | Required |
|-----------------|------|---------------------|-----------|----------|
| Eng Review | 1 | 2026-03-16 15:00 | CLEAR | YES |
| CEO Review | 0 | — | — | no |
| Design Review | 0 | — | — | no |
| Adversarial | 0 | — | — | no |
+--------------------------------------------------------------------+
| VERDICT: CLEARED — Eng Review passed |
+====================================================================+
```
**Review tiers:**
- **Eng Review (required by default):** The only review that gates shipping. Covers architecture, code quality, tests, performance. Can be disabled globally with \`gstack-config set skip_eng_review true\` (the "don't bother me" setting).
- **CEO Review (optional):** Use your judgment. Recommend it for big product/business changes, new user-facing features, or scope decisions. Skip for bug fixes, refactors, infra, and cleanup.
- **Design Review (optional):** Use your judgment. Recommend it for UI/UX changes. Skip for backend-only, infra, or prompt-only changes.
- **Adversarial Review (automatic):** Auto-scales by diff size. Small diffs (<50 lines) skip adversarial. Medium diffs (50–199) get cross-model adversarial. Large diffs (200+) get all 4 passes: Claude structured, Codex structured, Claude adversarial subagent, Codex adversarial. No configuration needed.
**Verdict logic:**
- **CLEARED**: Eng Review has >= 1 entry within 7 days with status "clean" (or \`skip_eng_review\` is \`true\`)
- **NOT CLEARED**: Eng Review missing, stale (>7 days), or has open issues
- CEO, Design, and Codex reviews are shown for context but never block shipping
- If \`skip_eng_review\` config is \`true\`, Eng Review shows "SKIPPED (global)" and verdict is CLEARED
**Staleness detection:** After displaying the dashboard, check if any existing reviews may be stale:
- Parse the \`---HEAD---\` section from the bash output to get the current HEAD commit hash
- For each review entry that has a \`commit\` field: compare it against the current HEAD. If different, count elapsed commits: \`git rev-list --count STORED_COMMIT..HEAD\`. Display: "Note: {skill} review from {date} may be stale — {N} commits since review"
- For entries without a \`commit\` field (legacy entries): display "Note: {skill} review from {date} has no commit tracking — consider re-running for accurate staleness detection"
- If all reviews match the current HEAD, do not display any staleness notes
## Plan File Review Report
After displaying the Review Readiness Dashboard in conversation output, also update the
**plan file** itself so review status is visible to anyone reading the plan.
### Detect the plan file
1. Check if there is an active plan file in this conversation (the host provides plan file
paths in system messages — look for plan file references in the conversation context).
2. If not found, skip this section silently — not every review runs in plan mode.
### Generate the report
Read the review log output you already have from the Review Readiness Dashboard step above.
Parse each JSONL entry. Each skill logs different fields:
- **plan-ceo-review**: \`status\`, \`unresolved\`, \`critical_gaps\`, \`mode\`, \`scope_proposed\`, \`scope_accepted\`, \`scope_deferred\`, \`commit\`
→ Findings: "{scope_proposed} proposals, {scope_accepted} accepted, {scope_deferred} deferred"
→ If scope fields are 0 or missing (HOLD/REDUCTION mode): "mode: {mode}, {critical_gaps} critical gaps"
- **plan-eng-review**: \`status\`, \`unresolved\`, \`critical_gaps\`, \`issues_found\`, \`mode\`, \`commit\`
→ Findings: "{issues_found} issues, {critical_gaps} critical gaps"
- **plan-design-review**: \`status\`, \`initial_score\`, \`overall_score\`, \`unresolved\`, \`decisions_made\`, \`commit\`
→ Findings: "score: {initial_score}/10 → {overall_score}/10, {decisions_made} decisions"
- **codex-review**: \`status\`, \`gate\`, \`findings\`, \`findings_fixed\`
→ Findings: "{findings} findings, {findings_fixed}/{findings} fixed"
All fields needed for the Findings column are now present in the JSONL entries.
For the review you just completed, you may use richer details from your own Completion
Summary. For prior reviews, use the JSONL fields directly — they contain all required data.
Produce this markdown table:
\`\`\`markdown
## GSTACK REVIEW REPORT
| Review | Trigger | Why | Runs | Status | Findings |
|--------|---------|-----|------|--------|----------|
| CEO Review | \`/plan-ceo-review\` | Scope & strategy | {runs} | {status} | {findings} |
| Codex Review | \`/codex review\` | Independent 2nd opinion | {runs} | {status} | {findings} |
| Eng Review | \`/plan-eng-review\` | Architecture & tests (required) | {runs} | {status} | {findings} |
| Design Review | \`/plan-design-review\` | UI/UX gaps | {runs} | {status} | {findings} |
\`\`\`
Below the table, add these lines (omit any that are empty/not applicable):
- **CODEX:** (only if codex-review ran) — one-line summary of codex fixes
- **CROSS-MODEL:** (only if both Claude and Codex reviews exist) — overlap analysis
- **UNRESOLVED:** total unresolved decisions across all reviews
- **VERDICT:** list reviews that are CLEAR (e.g., "CEO + ENG CLEARED — ready to implement").
If Eng Review is not CLEAR and not skipped globally, append "eng review required".
### Write to the plan file
**PLAN MODE EXCEPTION — ALWAYS RUN:** This writes to the plan file, which is the one
file you are allowed to edit in plan mode. The plan file review report is part of the
plan's living status.
- Search the plan file for a \`## GSTACK REVIEW REPORT\` section **anywhere** in the file
(not just at the end — content may have been added after it).
- If found, **replace it** entirely using the Edit tool. Match from \`## GSTACK REVIEW REPORT\`
through either the next \`## \` heading or end of file, whichever comes first. This ensures
content added after the report section is preserved, not eaten. If the Edit fails
(e.g., concurrent edit changed the content), re-read the plan file and retry once.
- If no such section exists, **append it** to the end of the plan file.
- Always place it as the very last section in the plan file. If it was found mid-file,
move it: delete the old location and append at the end.
## Next Steps — Review Chaining
After displaying the Review Readiness Dashboard, check if additional reviews would be valuable. Read the dashboard output to see which reviews have already been run and whether they are stale.
**Suggest /plan-design-review if UI changes exist and no design review has been run** — detect from the test diagram, architecture review, or any section that touched frontend components, CSS, views, or user-facing interaction flows. If an existing design review's commit hash shows it predates significant changes found in this eng review, note that it may be stale.
**Mention /plan-ceo-review if this is a significant product change and no CEO review exists** — this is a soft suggestion, not a push. CEO review is optional. Only mention it if the plan introduces new user-facing features, changes product direction, or expands scope substantially.
**Note staleness** of existing CEO or design reviews if this eng review found assumptions that contradict them, or if the commit hash shows significant drift.
**If no additional reviews are needed** (or `skip_eng_review` is `true` in the dashboard config, meaning this eng review was optional): state "All relevant reviews complete. Run /ship when ready."
Use AskUserQuestion with only the applicable options:
- **A)** Run /plan-design-review (only if UI scope detected and no design review exists)
- **B)** Run /plan-ceo-review (only if significant product change and no CEO review exists)
- **C)** Ready to implement — run /ship when done
## Unresolved decisions
If the user does not respond to an AskUserQuestion or interrupts to move on, note which decisions were left unresolved. At the end of the review, list these as "Unresolved decisions that may bite you later" — never silently default to an option.

311
skills/plan-eng-review/SKILL.md.tmpl
View File

@ -1,311 +0,0 @@
---
name: plan-eng-review
version: 1.0.0
description: |
Eng manager-mode plan review. Lock in the execution plan — architecture,
data flow, diagrams, edge cases, test coverage, performance. Walks through
issues interactively with opinionated recommendations. Use when asked to
"review the architecture", "engineering review", or "lock in the plan".
Proactively suggest when the user has a plan or design doc and is about to
start coding — to catch architecture issues before implementation.
benefits-from: [office-hours]
allowed-tools:
- Read
- Write
- Grep
- Glob
- AskUserQuestion
- Bash
- WebSearch
---
{{PREAMBLE}}
# Plan Review Mode
Review this plan thoroughly before making any code changes. For every issue or recommendation, explain the concrete tradeoffs, give me an opinionated recommendation, and ask for my input before assuming a direction.
## Priority hierarchy
If you are running low on context or the user asks you to compress: Step 0 > Test diagram > Opinionated recommendations > Everything else. Never skip Step 0 or the test diagram.
## My engineering preferences (use these to guide your recommendations):
* DRY is important—flag repetition aggressively.
* Well-tested code is non-negotiable; I'd rather have too many tests than too few.
* I want code that's "engineered enough" — not under-engineered (fragile, hacky) and not over-engineered (premature abstraction, unnecessary complexity).
* I err on the side of handling more edge cases, not fewer; thoughtfulness > speed.
* Bias toward explicit over clever.
* Minimal diff: achieve the goal with the fewest new abstractions and files touched.
## Cognitive Patterns — How Great Eng Managers Think
These are not additional checklist items. They are the instincts that experienced engineering leaders develop over years — the pattern recognition that separates "reviewed the code" from "caught the landmine." Apply them throughout your review.
1. **State diagnosis** — Teams exist in four states: falling behind, treading water, repaying debt, innovating. Each demands a different intervention (Larson, An Elegant Puzzle).
2. **Blast radius instinct** — Every decision evaluated through "what's the worst case and how many systems/people does it affect?"
3. **Boring by default** — "Every company gets about three innovation tokens." Everything else should be proven technology (McKinley, Choose Boring Technology).
4. **Incremental over revolutionary** — Strangler fig, not big bang. Canary, not global rollout. Refactor, not rewrite (Fowler).
5. **Systems over heroes** — Design for tired humans at 3am, not your best engineer on their best day.
6. **Reversibility preference** — Feature flags, A/B tests, incremental rollouts. Make the cost of being wrong low.
7. **Failure is information** — Blameless postmortems, error budgets, chaos engineering. Incidents are learning opportunities, not blame events (Allspaw, Google SRE).
8. **Org structure IS architecture** — Conway's Law in practice. Design both intentionally (Skelton/Pais, Team Topologies).
9. **DX is product quality** — Slow CI, bad local dev, painful deploys → worse software, higher attrition. Developer experience is a leading indicator.
10. **Essential vs accidental complexity** — Before adding anything: "Is this solving a real problem or one we created?" (Brooks, No Silver Bullet).
11. **Two-week smell test** — If a competent engineer can't ship a small feature in two weeks, you have an onboarding problem disguised as architecture.
12. **Glue work awareness** — Recognize invisible coordination work. Value it, but don't let people get stuck doing only glue (Reilly, The Staff Engineer's Path).
13. **Make the change easy, then make the easy change** — Refactor first, implement second. Never structural + behavioral changes simultaneously (Beck).
14. **Own your code in production** — No wall between dev and ops. "The DevOps movement is ending because there are only engineers who write code and own it in production" (Majors).
15. **Error budgets over uptime targets** — SLO of 99.9% = 0.1% downtime *budget to spend on shipping*. Reliability is resource allocation (Google SRE).
When evaluating architecture, think "boring by default." When reviewing tests, think "systems over heroes." When assessing complexity, ask Brooks's question. When a plan introduces new infrastructure, check whether it's spending an innovation token wisely.
## Documentation and diagrams:
* I value ASCII art diagrams highly — for data flow, state machines, dependency graphs, processing pipelines, and decision trees. Use them liberally in plans and design docs.
* For particularly complex designs or behaviors, embed ASCII diagrams directly in code comments in the appropriate places: Models (data relationships, state transitions), Controllers (request flow), Concerns (mixin behavior), Services (processing pipelines), and Tests (what's being set up and why) when the test structure is non-obvious.
* **Diagram maintenance is part of the change.** When modifying code that has ASCII diagrams in comments nearby, review whether those diagrams are still accurate. Update them as part of the same commit. Stale diagrams are worse than no diagrams — they actively mislead. Flag any stale diagrams you encounter during review even if they're outside the immediate scope of the change.
## BEFORE YOU START:
### Design Doc Check
```bash
SLUG=$(~/.claude/skills/gstack/browse/bin/remote-slug 2>/dev/null || basename "$(git rev-parse --show-toplevel 2>/dev/null || pwd)")
BRANCH=$(git rev-parse --abbrev-ref HEAD 2>/dev/null | tr '/' '-' || echo 'no-branch')
DESIGN=$(ls -t ~/.gstack/projects/$SLUG/*-$BRANCH-design-*.md 2>/dev/null | head -1)
[ -z "$DESIGN" ] && DESIGN=$(ls -t ~/.gstack/projects/$SLUG/*-design-*.md 2>/dev/null | head -1)
[ -n "$DESIGN" ] && echo "Design doc found: $DESIGN" || echo "No design doc found"
```
If a design doc exists, read it. Use it as the source of truth for the problem statement, constraints, and chosen approach. If it has a `Supersedes:` field, note that this is a revised design — check the prior version for context on what changed and why.
{{BENEFITS_FROM}}
### Step 0: Scope Challenge
Before reviewing anything, answer these questions:
1. **What existing code already partially or fully solves each sub-problem?** Can we capture outputs from existing flows rather than building parallel ones?
2. **What is the minimum set of changes that achieves the stated goal?** Flag any work that could be deferred without blocking the core objective. Be ruthless about scope creep.
3. **Complexity check:** If the plan touches more than 8 files or introduces more than 2 new classes/services, treat that as a smell and challenge whether the same goal can be achieved with fewer moving parts.
4. **Search check:** For each architectural pattern, infrastructure component, or concurrency approach the plan introduces:
- Does the runtime/framework have a built-in? Search: "{framework} {pattern} built-in"
- Is the chosen approach current best practice? Search: "{pattern} best practice {current year}"
- Are there known footguns? Search: "{framework} {pattern} pitfalls"
If WebSearch is unavailable, skip this check and note: "Search unavailable — proceeding with in-distribution knowledge only."
If the plan rolls a custom solution where a built-in exists, flag it as a scope reduction opportunity. Annotate recommendations with **[Layer 1]**, **[Layer 2]**, **[Layer 3]**, or **[EUREKA]** (see preamble's Search Before Building section). If you find a eureka moment — a reason the standard approach is wrong for this case — present it as an architectural insight.
5. **TODOS cross-reference:** Read `TODOS.md` if it exists. Are any deferred items blocking this plan? Can any deferred items be bundled into this PR without expanding scope? Does this plan create new work that should be captured as a TODO?
5. **Completeness check:** Is the plan doing the complete version or a shortcut? With AI-assisted coding, the cost of completeness (100% test coverage, full edge case handling, complete error paths) is 10-100x cheaper than with a human team. If the plan proposes a shortcut that saves human-hours but only saves minutes with CC+gstack, recommend the complete version. Boil the lake.
If the complexity check triggers (8+ files or 2+ new classes/services), proactively recommend scope reduction via AskUserQuestion — explain what's overbuilt, propose a minimal version that achieves the core goal, and ask whether to reduce or proceed as-is. If the complexity check does not trigger, present your Step 0 findings and proceed directly to Section 1.
### Step 0.5: Codex plan review (optional)
Check if the Codex CLI is available: `which codex 2>/dev/null`
If available, after presenting Step 0 findings, use AskUserQuestion:
```
Want an independent Codex (OpenAI) review of this plan before the detailed review?
A) Yes — let Codex critique the plan independently
B) No — proceed with the Claude review only
```
If the user chooses A: tell Codex to read the plan file itself (avoids ARG_MAX limits for large plans):
```bash
codex exec "You are a brutally honest technical reviewer. Read the plan file at <plan-file-path> and review it for: logical gaps and unstated assumptions, missing error handling or edge cases, overcomplexity (is there a simpler approach?), feasibility risks (what could go wrong?), and missing dependencies or sequencing issues. Be direct. Be terse. No compliments. Just the problems." -s read-only -c 'model_reasoning_effort="high"' --enable web_search_cached
```
Replace `<plan-file-path>` with the actual path to the plan file detected earlier. Codex has filesystem access in read-only mode and will read the file itself.
Present the full output under a `CODEX SAYS (plan review):` header. Note any concerns
that should inform the subsequent engineering review sections.
If Codex is not available, skip silently.
Always work through the full interactive review: one section at a time (Architecture → Code Quality → Tests → Performance) with at most 8 top issues per section.
**Critical: Once the user accepts or rejects a scope reduction recommendation, commit fully.** Do not re-argue for smaller scope during later review sections. Do not silently reduce scope or skip planned components.
## Review Sections (after scope is agreed)
### 1. Architecture review
Evaluate:
* Overall system design and component boundaries.
* Dependency graph and coupling concerns.
* Data flow patterns and potential bottlenecks.
* Scaling characteristics and single points of failure.
* Security architecture (auth, data access, API boundaries).
* Whether key flows deserve ASCII diagrams in the plan or in code comments.
* For each new codepath or integration point, describe one realistic production failure scenario and whether the plan accounts for it.
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
### 2. Code quality review
Evaluate:
* Code organization and module structure.
* DRY violations—be aggressive here.
* Error handling patterns and missing edge cases (call these out explicitly).
* Technical debt hotspots.
* Areas that are over-engineered or under-engineered relative to my preferences.
* Existing ASCII diagrams in touched files — are they still accurate after this change?
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
### 3. Test review
Make a diagram of all new UX, new data flow, new codepaths, and new branching if statements or outcomes. For each, note what is new about the features discussed in this branch and plan. Then, for each new item in the diagram, make sure there is a corresponding test.
For LLM/prompt changes: check the "Prompt/LLM changes" file patterns listed in CLAUDE.md. If this plan touches ANY of those patterns, state which eval suites must be run, which cases should be added, and what baselines to compare against. Then use AskUserQuestion to confirm the eval scope with the user.
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
### Test Plan Artifact
After producing the test diagram, write a test plan artifact to the project directory so `/qa` and `/qa-only` can consume it as primary test input (replacing the lossy git-diff heuristic):
```bash
source <(~/.claude/skills/gstack/bin/gstack-slug 2>/dev/null) && mkdir -p ~/.gstack/projects/$SLUG
USER=$(whoami)
DATETIME=$(date +%Y%m%d-%H%M%S)
```
Write to `~/.gstack/projects/{slug}/{user}-{branch}-test-plan-{datetime}.md`:
```markdown
# Test Plan
Generated by /plan-eng-review on {date}
Branch: {branch}
Repo: {owner/repo}
## Affected Pages/Routes
- {URL path} — {what to test and why}
## Key Interactions to Verify
- {interaction description} on {page}
## Edge Cases
- {edge case} on {page}
## Critical Paths
- {end-to-end flow that must work}
```
This file is consumed by `/qa` and `/qa-only` as primary test input. Include only the information that helps a QA tester know **what to test and where** — not implementation details.
### 4. Performance review
Evaluate:
* N+1 queries and database access patterns.
* Memory-usage concerns.
* Caching opportunities.
* Slow or high-complexity code paths.
**STOP.** For each issue found in this section, call AskUserQuestion individually. One issue per call. Present options, state your recommendation, explain WHY. Do NOT batch multiple issues into one AskUserQuestion. Only proceed to the next section after ALL issues in this section are resolved.
## CRITICAL RULE — How to ask questions
Follow the AskUserQuestion format from the Preamble above. Additional rules for plan reviews:
* **One issue = one AskUserQuestion call.** Never combine multiple issues into one question.
* Describe the problem concretely, with file and line references.
* Present 2-3 options, including "do nothing" where that's reasonable.
* For each option, specify in one line: effort (human: ~X / CC: ~Y), risk, and maintenance burden. If the complete option is only marginally more effort than the shortcut with CC, recommend the complete option.
* **Map the reasoning to my engineering preferences above.** One sentence connecting your recommendation to a specific preference (DRY, explicit > clever, minimal diff, etc.).
* Label with issue NUMBER + option LETTER (e.g., "3A", "3B").
* **Escape hatch:** If a section has no issues, say so and move on. If an issue has an obvious fix with no real alternatives, state what you'll do and move on — don't waste a question on it. Only use AskUserQuestion when there is a genuine decision with meaningful tradeoffs.
## Required outputs
### "NOT in scope" section
Every plan review MUST produce a "NOT in scope" section listing work that was considered and explicitly deferred, with a one-line rationale for each item.
### "What already exists" section
List existing code/flows that already partially solve sub-problems in this plan, and whether the plan reuses them or unnecessarily rebuilds them.
### TODOS.md updates
After all review sections are complete, present each potential TODO as its own individual AskUserQuestion. Never batch TODOs — one per question. Never silently skip this step. Follow the format in `.claude/skills/review/TODOS-format.md`.
For each TODO, describe:
* **What:** One-line description of the work.
* **Why:** The concrete problem it solves or value it unlocks.
* **Pros:** What you gain by doing this work.
* **Cons:** Cost, complexity, or risks of doing it.
* **Context:** Enough detail that someone picking this up in 3 months understands the motivation, the current state, and where to start.
* **Depends on / blocked by:** Any prerequisites or ordering constraints.
Then present options: **A)** Add to TODOS.md **B)** Skip — not valuable enough **C)** Build it now in this PR instead of deferring.
Do NOT just append vague bullet points. A TODO without context is worse than no TODO — it creates false confidence that the idea was captured while actually losing the reasoning.
### Diagrams
The plan itself should use ASCII diagrams for any non-trivial data flow, state machine, or processing pipeline. Additionally, identify which files in the implementation should get inline ASCII diagram comments — particularly Models with complex state transitions, Services with multi-step pipelines, and Concerns with non-obvious mixin behavior.
### Failure modes
For each new codepath identified in the test review diagram, list one realistic way it could fail in production (timeout, nil reference, race condition, stale data, etc.) and whether:
1. A test covers that failure
2. Error handling exists for it
3. The user would see a clear error or a silent failure
If any failure mode has no test AND no error handling AND would be silent, flag it as a **critical gap**.
### Completion summary
At the end of the review, fill in and display this summary so the user can see all findings at a glance:
- Step 0: Scope Challenge — ___ (scope accepted as-is / scope reduced per recommendation)
- Architecture Review: ___ issues found
- Code Quality Review: ___ issues found
- Test Review: diagram produced, ___ gaps identified
- Performance Review: ___ issues found
- NOT in scope: written
- What already exists: written
- TODOS.md updates: ___ items proposed to user
- Failure modes: ___ critical gaps flagged
- Lake Score: X/Y recommendations chose complete option
## Retrospective learning
Check the git log for this branch. If there are prior commits suggesting a previous review cycle (e.g., review-driven refactors, reverted changes), note what was changed and whether the current plan touches the same areas. Be more aggressive reviewing areas that were previously problematic.
## Formatting rules
* NUMBER issues (1, 2, 3...) and LETTERS for options (A, B, C...).
* Label with NUMBER + LETTER (e.g., "3A", "3B").
* One sentence max per option. Pick in under 5 seconds.
* After each review section, pause and ask for feedback before moving on.
## Review Log
After producing the Completion Summary above, persist the review result.
**PLAN MODE EXCEPTION — ALWAYS RUN:** This command writes review metadata to
`~/.gstack/` (user config directory, not project files). The skill preamble
already writes to `~/.gstack/sessions/` and `~/.gstack/analytics/` — this is
the same pattern. The review dashboard depends on this data. Skipping this
command breaks the review readiness dashboard in /ship.
```bash
~/.claude/skills/gstack/bin/gstack-review-log '{"skill":"plan-eng-review","timestamp":"TIMESTAMP","status":"STATUS","unresolved":N,"critical_gaps":N,"issues_found":N,"mode":"MODE","commit":"COMMIT"}'
```
Substitute values from the Completion Summary:
- **TIMESTAMP**: current ISO 8601 datetime
- **STATUS**: "clean" if 0 unresolved decisions AND 0 critical gaps; otherwise "issues_open"
- **unresolved**: number from "Unresolved decisions" count
- **critical_gaps**: number from "Failure modes: ___ critical gaps flagged"
- **issues_found**: total issues found across all review sections (Architecture + Code Quality + Performance + Test gaps)
- **MODE**: FULL_REVIEW / SCOPE_REDUCED
- **COMMIT**: output of `git rev-parse --short HEAD`
{{REVIEW_DASHBOARD}}
{{PLAN_FILE_REVIEW_REPORT}}
## Next Steps — Review Chaining
After displaying the Review Readiness Dashboard, check if additional reviews would be valuable. Read the dashboard output to see which reviews have already been run and whether they are stale.
**Suggest /plan-design-review if UI changes exist and no design review has been run** — detect from the test diagram, architecture review, or any section that touched frontend components, CSS, views, or user-facing interaction flows. If an existing design review's commit hash shows it predates significant changes found in this eng review, note that it may be stale.
**Mention /plan-ceo-review if this is a significant product change and no CEO review exists** — this is a soft suggestion, not a push. CEO review is optional. Only mention it if the plan introduces new user-facing features, changes product direction, or expands scope substantially.
**Note staleness** of existing CEO or design reviews if this eng review found assumptions that contradict them, or if the commit hash shows significant drift.
**If no additional reviews are needed** (or `skip_eng_review` is `true` in the dashboard config, meaning this eng review was optional): state "All relevant reviews complete. Run /ship when ready."
Use AskUserQuestion with only the applicable options:
- **A)** Run /plan-design-review (only if UI scope detected and no design review exists)
- **B)** Run /plan-ceo-review (only if significant product change and no CEO review exists)
- **C)** Ready to implement — run /ship when done
## Unresolved decisions
If the user does not respond to an AskUserQuestion or interrupts to move on, note which decisions were left unresolved. At the end of the review, list these as "Unresolved decisions that may bite you later" — never silently default to an option.

82
skills/rust-engineer/SKILL.md
View File

@ -1,82 +0,0 @@
---
name: rust-engineer
description: >
Rust 系统编程专家。当用户需要 Rust 所有权/借用/生命周期、async/await、宏开发、trait 设计、错误处理 Result/Option、性能优化、WebAssembly/WASM，或说 "Rust"、"所有权"、"借用检查" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, Bash
maturity: imported
last-reviewed: 2026-03-03
composable: true
enhances: [performance-expert, edge-computing-expert]
---
# Rust Engineer
Senior Rust engineer with deep expertise in Rust 2021 edition, systems programming, memory safety, and zero-cost abstractions. Specializes in building reliable, high-performance software leveraging Rust's ownership system.
## Role Definition
You are a senior Rust engineer with 10+ years of systems programming experience. You specialize in Rust's ownership model, async programming with tokio, trait-based design, and performance optimization. You build memory-safe, concurrent systems with zero-cost abstractions.
## When to Use This Skill
- Building systems-level applications in Rust
- Implementing ownership and borrowing patterns
- Designing trait hierarchies and generic APIs
- Setting up async/await with tokio or async-std
- Optimizing for performance and memory safety
- Creating FFI bindings and unsafe abstractions
## Core Workflow
1. **Analyze ownership** - Design lifetime relationships and borrowing patterns
2. **Design traits** - Create trait hierarchies with generics and associated types
3. **Implement safely** - Write idiomatic Rust with minimal unsafe code
4. **Handle errors** - Use Result/Option with ? operator and custom error types
5. **Test thoroughly** - Unit tests, integration tests, property testing, benchmarks
## Reference Guide
Load detailed guidance based on context:
| Topic | Reference | Load When |
|-------|-----------|-----------|
| Ownership | `references/ownership.md` | Lifetimes, borrowing, smart pointers, Pin |
| Traits | `references/traits.md` | Trait design, generics, associated types, derive |
| Error Handling | `references/error-handling.md` | Result, Option, ?, custom errors, thiserror |
| Async | `references/async.md` | async/await, tokio, futures, streams, concurrency |
| Testing | `references/testing.md` | Unit/integration tests, proptest, benchmarks |
## Constraints
### MUST DO
- Use ownership and borrowing for memory safety
- Minimize unsafe code (document all unsafe blocks)
- Use type system for compile-time guarantees
- Handle all errors explicitly (Result/Option)
- Add comprehensive documentation with examples
- Run clippy and fix all warnings
- Use cargo fmt for consistent formatting
- Write tests including doctests
### MUST NOT DO
- Use unwrap() in production code (prefer expect() with messages)
- Create memory leaks or dangling pointers
- Use unsafe without documenting safety invariants
- Ignore clippy warnings
- Mix blocking and async code incorrectly
- Skip error handling
- Use String when &str suffices
- Clone unnecessarily (use borrowing)
## Output Templates
When implementing Rust features, provide:
1. Type definitions (structs, enums, traits)
2. Implementation with proper ownership
3. Error handling with custom error types
4. Tests (unit, integration, doctests)
5. Brief explanation of design decisions
## Knowledge Reference
Rust 2021, Cargo, ownership/borrowing, lifetimes, traits, generics, async/await, tokio, Result/Option, thiserror/anyhow, serde, clippy, rustfmt, cargo-test, criterion benchmarks, MIRI, unsafe Rust

458
skills/rust-engineer/references/async.md
View File

@ -1,458 +0,0 @@
# Async Programming in Rust
## Basic Async/Await
```rust
use tokio;
// Async function returns a Future
async fn fetch_data(url: &str) -> Result<String, reqwest::Error> {
let response = reqwest::get(url).await?;
let body = response.text().await?;
Ok(body)
}
// Tokio runtime
#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
let data = fetch_data("https://api.example.com").await?;
println!("Data: {}", data);
Ok(())
}
// Manual runtime creation
fn main() {
let runtime = tokio::runtime::Runtime::new().unwrap();
runtime.block_on(async {
println!("Hello from async context");
});
}
```
## Concurrent Execution
```rust
use tokio;
// Sequential execution
async fn sequential() {
let result1 = async_operation1().await;
let result2 = async_operation2().await; // Waits for operation1
}
// Concurrent execution with join!
async fn concurrent() {
let (result1, result2) = tokio::join!(
async_operation1(),
async_operation2()
);
}
// Concurrent with try_join! (stops on first error)
async fn concurrent_with_errors() -> Result<(), Box<dyn std::error::Error>> {
let (result1, result2) = tokio::try_join!(
fallible_operation1(),
fallible_operation2()
)?;
Ok(())
}
// Spawning tasks
async fn spawn_tasks() {
let handle1 = tokio::spawn(async {
// This runs on a separate task
expensive_computation().await
});
let handle2 = tokio::spawn(async {
another_computation().await
});
// Wait for both to complete
let result1 = handle1.await.unwrap();
let result2 = handle2.await.unwrap();
}
```
## Select and Race Conditions
```rust
use tokio::time::{sleep, Duration};
// select! - wait for first to complete
async fn first_to_complete() {
tokio::select! {
result = async_operation1() => {
println!("Operation 1 completed first: {:?}", result);
}
result = async_operation2() => {
println!("Operation 2 completed first: {:?}", result);
}
}
}
// Timeout pattern
async fn with_timeout() -> Result<String, &'static str> {
tokio::select! {
result = fetch_data("https://api.example.com") => {
result.map_err(|_| "Fetch failed")
}
_ = sleep(Duration::from_secs(5)) => {
Err("Timeout")
}
}
}
// Cancellation with select!
async fn cancellable_operation(mut cancel_rx: tokio::sync::watch::Receiver<bool>) {
tokio::select! {
result = long_running_task() => {
println!("Task completed: {:?}", result);
}
_ = cancel_rx.changed() => {
println!("Task cancelled");
}
}
}
```
## Streams
```rust
use tokio_stream::{self as stream, StreamExt};
// Creating streams
async fn stream_example() {
let mut stream = stream::iter(vec![1, 2, 3, 4, 5]);
while let Some(value) = stream.next().await {
println!("Value: {}", value);
}
}
// Stream combinators
async fn stream_combinators() {
let stream = stream::iter(vec![1, 2, 3, 4, 5])
.filter(|x| *x % 2 == 0)
.map(|x| x * 2);
let results: Vec<_> = stream.collect().await;
println!("Results: {:?}", results);
}
// Async stream processing
use futures::stream::{self, StreamExt};
async fn process_stream() {
let stream = stream::iter(vec![1, 2, 3, 4, 5])
.then(|x| async move {
tokio::time::sleep(Duration::from_millis(100)).await;
x * 2
});
stream.for_each(|x| async move {
println!("Processed: {}", x);
}).await;
}
```
## Channels for Communication
```rust
use tokio::sync::{mpsc, oneshot, broadcast, watch};
// mpsc: multiple producer, single consumer
async fn mpsc_example() {
let (tx, mut rx) = mpsc::channel(32);
tokio::spawn(async move {
tx.send("Hello").await.unwrap();
tx.send("World").await.unwrap();
});
while let Some(msg) = rx.recv().await {
println!("Received: {}", msg);
}
}
// oneshot: single value, one-time use
async fn oneshot_example() {
let (tx, rx) = oneshot::channel();
tokio::spawn(async move {
tx.send("Result").unwrap();
});
let result = rx.await.unwrap();
println!("Got: {}", result);
}
// broadcast: multiple producers, multiple consumers
async fn broadcast_example() {
let (tx, mut rx1) = broadcast::channel(16);
let mut rx2 = tx.subscribe();
tokio::spawn(async move {
tx.send("Message").unwrap();
});
println!("rx1: {}", rx1.recv().await.unwrap());
println!("rx2: {}", rx2.recv().await.unwrap());
}
// watch: single producer, multiple consumers (last value)
async fn watch_example() {
let (tx, mut rx) = watch::channel("initial");
tokio::spawn(async move {
loop {
rx.changed().await.unwrap();
println!("Value changed to: {}", *rx.borrow());
}
});
tx.send("updated").unwrap();
}
```
## Shared State
```rust
use std::sync::Arc;
use tokio::sync::{Mutex, RwLock};
// Mutex for exclusive access
async fn mutex_example() {
let data = Arc::new(Mutex::new(0));
let mut handles = vec![];
for _ in 0..10 {
let data = Arc::clone(&data);
let handle = tokio::spawn(async move {
let mut lock = data.lock().await;
*lock += 1;
});
handles.push(handle);
}
for handle in handles {
handle.await.unwrap();
}
println!("Final value: {}", *data.lock().await);
}
// RwLock for read-write patterns
async fn rwlock_example() {
let data = Arc::new(RwLock::new(vec![1, 2, 3]));
// Multiple readers
let data1 = Arc::clone(&data);
tokio::spawn(async move {
let read = data1.read().await;
println!("Read: {:?}", *read);
});
let data2 = Arc::clone(&data);
tokio::spawn(async move {
let read = data2.read().await;
println!("Read: {:?}", *read);
});
// Single writer
tokio::time::sleep(Duration::from_millis(100)).await;
let mut write = data.write().await;
write.push(4);
}
```
## Async Traits (with async-trait)
```rust
use async_trait::async_trait;
#[async_trait]
trait AsyncRepository {
async fn find_by_id(&self, id: u64) -> Result<User, Error>;
async fn save(&self, user: User) -> Result<(), Error>;
}
struct DatabaseRepository {
pool: sqlx::PgPool,
}
#[async_trait]
impl AsyncRepository for DatabaseRepository {
async fn find_by_id(&self, id: u64) -> Result<User, Error> {
sqlx::query_as("SELECT * FROM users WHERE id = $1")
.bind(id)
.fetch_one(&self.pool)
.await
.map_err(Into::into)
}
async fn save(&self, user: User) -> Result<(), Error> {
sqlx::query("INSERT INTO users (name, email) VALUES ($1, $2)")
.bind(&user.name)
.bind(&user.email)
.execute(&self.pool)
.await?;
Ok(())
}
}
```
## Pin and Futures
```rust
use std::pin::Pin;
use std::future::Future;
use std::task::{Context, Poll};
// Manual Future implementation
struct DelayedValue {
value: i32,
delay: tokio::time::Sleep,
}
impl Future for DelayedValue {
type Output = i32;
fn poll(mut self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<Self::Output> {
match Pin::new(&mut self.delay).poll(cx) {
Poll::Ready(_) => Poll::Ready(self.value),
Poll::Pending => Poll::Pending,
}
}
}
// Using pinned futures
async fn use_pinned() {
let future = DelayedValue {
value: 42,
delay: tokio::time::sleep(Duration::from_secs(1)),
};
let result = future.await;
println!("Result: {}", result);
}
```
## Background Tasks and Graceful Shutdown
```rust
use tokio::signal;
async fn background_task(mut shutdown: tokio::sync::watch::Receiver<bool>) {
loop {
tokio::select! {
_ = tokio::time::sleep(Duration::from_secs(1)) => {
println!("Background task running...");
}
_ = shutdown.changed() => {
println!("Shutting down background task");
break;
}
}
}
}
#[tokio::main]
async fn main() {
let (shutdown_tx, shutdown_rx) = tokio::sync::watch::channel(false);
let task = tokio::spawn(background_task(shutdown_rx));
// Wait for ctrl-c
signal::ctrl_c().await.unwrap();
println!("Received shutdown signal");
// Signal shutdown
shutdown_tx.send(true).unwrap();
// Wait for task to complete
task.await.unwrap();
}
```
## Error Handling in Async
```rust
use thiserror::Error;
#[derive(Error, Debug)]
enum AsyncError {
#[error("Network error: {0}")]
Network(#[from] reqwest::Error),
#[error("Timeout")]
Timeout,
#[error("Task failed")]
TaskFailed(#[from] tokio::task::JoinError),
}
async fn robust_operation() -> Result<String, AsyncError> {
let timeout = Duration::from_secs(5);
let result = tokio::time::timeout(timeout, async {
reqwest::get("https://api.example.com")
.await?
.text()
.await
})
.await
.map_err(|_| AsyncError::Timeout)??;
Ok(result)
}
```
## Runtime Configuration
```rust
// Custom runtime configuration
fn main() {
let runtime = tokio::runtime::Builder::new_multi_thread()
.worker_threads(4)
.thread_name("my-worker")
.thread_stack_size(3 * 1024 * 1024)
.enable_all()
.build()
.unwrap();
runtime.block_on(async {
println!("Running on custom runtime");
});
}
// Current-thread runtime (single-threaded)
fn single_threaded() {
let runtime = tokio::runtime::Builder::new_current_thread()
.enable_all()
.build()
.unwrap();
runtime.block_on(async {
println!("Single-threaded async");
});
}
```
## Best Practices
- Use tokio::spawn for CPU-bound tasks on multi-threaded runtime
- Use spawn_blocking for blocking operations (file I/O, sync code)
- Prefer tokio::sync primitives over std::sync in async code
- Use channels for task communication instead of shared state when possible
- Always handle JoinHandle results (tasks can panic)
- Use select! for cancellation patterns
- Avoid holding locks across .await points
- Use timeout for all external I/O operations
- Implement graceful shutdown with channels
- Use async-trait for trait-based async code
- Prefer try_join! over manual error handling
- Use Arc<Mutex<T>> sparingly (channels often better)
- Test async code with tokio::test macro
- Monitor task spawning to prevent unbounded growth

334
skills/rust-engineer/references/error-handling.md
View File

@ -1,334 +0,0 @@
# Error Handling in Rust
## Result and Option Basics
```rust
// Result: operation that can fail
fn divide(a: f64, b: f64) -> Result<f64, String> {
if b == 0.0 {
Err("Division by zero".to_string())
} else {
Ok(a / b)
}
}
// Option: value that might be absent
fn find_user(id: u64) -> Option<User> {
if id == 1 {
Some(User { id, name: "Alice".to_string() })
} else {
None
}
}
// Using ? operator for propagation
fn calculate(a: f64, b: f64, c: f64) -> Result<f64, String> {
let x = divide(a, b)?; // Returns Err early if division fails
let y = divide(x, c)?;
Ok(y)
}
```
## Custom Error Types
```rust
use std::fmt;
// Manual error type
#[derive(Debug)]
enum AppError {
NotFound(String),
InvalidInput(String),
DatabaseError(String),
}
impl fmt::Display for AppError {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
match self {
AppError::NotFound(msg) => write!(f, "Not found: {}", msg),
AppError::InvalidInput(msg) => write!(f, "Invalid input: {}", msg),
AppError::DatabaseError(msg) => write!(f, "Database error: {}", msg),
}
}
}
impl std::error::Error for AppError {}
// Usage
fn get_user(id: u64) -> Result<User, AppError> {
if id == 0 {
return Err(AppError::InvalidInput("ID cannot be zero".to_string()));
}
// ... fetch user
Err(AppError::NotFound(format!("User {} not found", id)))
}
```
## Using thiserror
```rust
use thiserror::Error;
#[derive(Error, Debug)]
enum DataError {
#[error("Data not found: {0}")]
NotFound(String),
#[error("Invalid ID: {id}, reason: {reason}")]
InvalidId { id: u64, reason: String },
#[error("IO error")]
Io(#[from] std::io::Error),
#[error("Parse error")]
Parse(#[from] std::num::ParseIntError),
#[error("Database error: {0}")]
Database(#[from] sqlx::Error),
}
// Usage with automatic conversions
fn read_config(path: &str) -> Result<Config, DataError> {
let content = std::fs::read_to_string(path)?; // Auto-converts io::Error
let port: u16 = content.parse()?; // Auto-converts ParseIntError
Ok(Config { port })
}
```
## Using anyhow for Applications
```rust
use anyhow::{Result, Context, bail, ensure};
// Simple error handling for applications
fn process_file(path: &str) -> Result<()> {
let content = std::fs::read_to_string(path)
.context(format!("Failed to read file: {}", path))?;
ensure!(!content.is_empty(), "File is empty");
if content.len() > 1000 {
bail!("File too large");
}
// Process content...
Ok(())
}
// Adding context to errors
fn main() -> Result<()> {
process_file("config.txt")
.context("Failed to process configuration")?;
Ok(())
}
```
## Option Combinators
```rust
// map: transform Option<T> to Option<U>
let num: Option<i32> = Some(5);
let doubled = num.map(|n| n * 2); // Some(10)
// and_then: chain operations
let result = Some(5)
.and_then(|n| if n > 0 { Some(n * 2) } else { None })
.and_then(|n| Some(n + 1)); // Some(11)
// or: provide alternative
let value = None.or(Some(42)); // Some(42)
// unwrap_or: provide default
let value = None.unwrap_or(42); // 42
// unwrap_or_else: compute default lazily
let value = None.unwrap_or_else(|| expensive_computation());
// filter: conditional None
let num = Some(5).filter(|&n| n > 10); // None
// Pattern matching
match find_user(1) {
Some(user) => println!("Found: {}", user.name),
None => println!("User not found"),
}
// if let for simple cases
if let Some(user) = find_user(1) {
println!("Found: {}", user.name);
}
```
## Result Combinators
```rust
// map: transform Ok value
let result: Result<i32, String> = Ok(5);
let doubled = result.map(|n| n * 2); // Ok(10)
// map_err: transform error
let result: Result<i32, &str> = Err("error");
let mapped = result.map_err(|e| e.to_uppercase()); // Err("ERROR")
// and_then: chain fallible operations
fn parse_then_double(s: &str) -> Result<i32, std::num::ParseIntError> {
s.parse::<i32>()
.and_then(|n| Ok(n * 2))
}
// or_else: provide alternative computation
let result = Err("error").or_else(|_| Ok(42)); // Ok(42)
// unwrap_or: provide default
let value = Err("error").unwrap_or(42); // 42
// expect: unwrap with custom panic message
let value = result.expect("Failed to parse number");
// Pattern matching
match divide(10.0, 2.0) {
Ok(result) => println!("Result: {}", result),
Err(e) => eprintln!("Error: {}", e),
}
```
## Error Conversion and From Trait
```rust
use std::io;
use std::num::ParseIntError;
#[derive(Debug)]
enum MyError {
Io(io::Error),
Parse(ParseIntError),
}
impl From<io::Error> for MyError {
fn from(err: io::Error) -> Self {
MyError::Io(err)
}
}
impl From<ParseIntError> for MyError {
fn from(err: ParseIntError) -> Self {
MyError::Parse(err)
}
}
// Now ? operator works with automatic conversion
fn read_and_parse(path: &str) -> Result<i32, MyError> {
let content = std::fs::read_to_string(path)?; // io::Error -> MyError
let number = content.trim().parse()?; // ParseIntError -> MyError
Ok(number)
}
```
## Advanced Error Patterns
```rust
// Multiple error sources with Box<dyn Error>
use std::error::Error;
fn complex_operation() -> Result<String, Box<dyn Error>> {
let file = std::fs::read_to_string("data.txt")?;
let number: i32 = file.trim().parse()?;
Ok(format!("Number: {}", number))
}
// Error with backtrace (nightly)
#[derive(Debug)]
struct DetailedError {
message: String,
backtrace: std::backtrace::Backtrace,
}
impl DetailedError {
fn new(message: impl Into<String>) -> Self {
Self {
message: message.into(),
backtrace: std::backtrace::Backtrace::capture(),
}
}
}
// Recoverable vs unrecoverable errors
fn might_fail(value: i32) -> Result<i32, String> {
if value < 0 {
Err("Negative value".to_string()) // Recoverable
} else if value > 1000 {
panic!("Value too large!"); // Unrecoverable
} else {
Ok(value * 2)
}
}
```
## Try Blocks (Nightly)
```rust
#![feature(try_blocks)]
// Try block for localized error handling
let result: Result<i32, Box<dyn Error>> = try {
let file = std::fs::read_to_string("config.txt")?;
let num: i32 = file.trim().parse()?;
num * 2
};
```
## Error Context Pattern
```rust
use thiserror::Error;
#[derive(Error, Debug)]
#[error("{message}")]
struct ContextError {
message: String,
#[source]
source: Option<Box<dyn Error + Send + Sync>>,
}
impl ContextError {
fn new(message: impl Into<String>) -> Self {
Self {
message: message.into(),
source: None,
}
}
fn with_source(mut self, source: impl Error + Send + Sync + 'static) -> Self {
self.source = Some(Box::new(source));
self
}
}
// Extension trait for adding context
trait Context<T> {
fn context(self, message: impl Into<String>) -> Result<T, ContextError>;
}
impl<T, E: Error + Send + Sync + 'static> Context<T> for Result<T, E> {
fn context(self, message: impl Into<String>) -> Result<T, ContextError> {
self.map_err(|e| ContextError::new(message).with_source(e))
}
}
```
## Best Practices
- Use Result for recoverable errors, panic! for unrecoverable bugs
- Prefer ? operator over unwrap() in production code
- Use expect() with descriptive messages instead of unwrap()
- Use thiserror for libraries (structured errors)
- Use anyhow for applications (simple error handling)
- Implement std::error::Error trait for custom error types
- Add context to errors as they propagate up the stack
- Use #[from] in thiserror for automatic conversions
- Document error conditions in function documentation
- Use Option::ok_or() to convert Option to Result
- Use Result::ok() to convert Result to Option (discarding error)
- Avoid String as error type (use custom types instead)
- Use ensure! and bail! from anyhow for cleaner checks
- Log errors at boundaries, return them in library code

278
skills/rust-engineer/references/ownership.md
View File

@ -1,278 +0,0 @@
# Ownership, Borrowing, and Lifetimes
## Ownership Patterns
```rust
// Move semantics (ownership transfer)
fn take_ownership(s: String) {
println!("{}", s);
} // s dropped here
// Borrowing (immutable reference)
fn borrow(s: &String) {
println!("{}", s);
} // s NOT dropped, caller still owns
// Mutable borrowing
fn borrow_mut(s: &mut String) {
s.push_str(" world");
}
// Usage
let s = String::from("hello");
borrow(&s); // OK, immutable borrow
let mut s2 = s; // Move, s no longer valid
borrow_mut(&mut s2); // OK, mutable borrow
```
## Lifetime Annotations
```rust
// Explicit lifetime: returned reference lives as long as input
fn longest<'a>(x: &'a str, y: &'a str) -> &'a str {
if x.len() > y.len() { x } else { y }
}
// Multiple lifetimes
fn first_word<'a, 'b>(s: &'a str, _other: &'b str) -> &'a str {
s.split_whitespace().next().unwrap_or("")
}
// Lifetime in structs
struct Excerpt<'a> {
part: &'a str,
}
impl<'a> Excerpt<'a> {
fn announce_and_return(&self, announcement: &str) -> &'a str {
println!("Attention: {}", announcement);
self.part
}
}
// Static lifetime (lives for entire program)
const GREETING: &'static str = "Hello, world!";
```
## Smart Pointers
```rust
use std::rc::Rc;
use std::cell::RefCell;
use std::sync::{Arc, Mutex};
// Box: heap allocation, single owner
let b = Box::new(5);
// Rc: reference counting (single-threaded)
let rc1 = Rc::new(vec![1, 2, 3]);
let rc2 = Rc::clone(&rc1); // Increment count
println!("Count: {}", Rc::strong_count(&rc1)); // 2
// Arc: atomic reference counting (thread-safe)
let arc1 = Arc::new(vec![1, 2, 3]);
let arc2 = Arc::clone(&arc1);
std::thread::spawn(move || {
println!("{:?}", arc2);
});
// RefCell: interior mutability (runtime borrow checking)
let data = RefCell::new(5);
*data.borrow_mut() += 1; // Mutable borrow at runtime
// Combining Rc + RefCell for shared mutable state
let shared = Rc::new(RefCell::new(vec![1, 2, 3]));
shared.borrow_mut().push(4);
// Combining Arc + Mutex for thread-safe shared state
let counter = Arc::new(Mutex::new(0));
let counter_clone = Arc::clone(&counter);
std::thread::spawn(move || {
let mut num = counter_clone.lock().unwrap();
*num += 1;
});
```
## Interior Mutability
```rust
use std::cell::{Cell, RefCell};
// Cell: Copy types only
let c = Cell::new(5);
c.set(10);
let val = c.get();
// RefCell: runtime borrow checking
let data = RefCell::new(vec![1, 2, 3]);
data.borrow_mut().push(4);
// Pattern: mock objects with interior mutability
struct MockLogger {
messages: RefCell<Vec<String>>,
}
impl MockLogger {
fn new() -> Self {
Self { messages: RefCell::new(Vec::new()) }
}
fn log(&self, msg: &str) {
self.messages.borrow_mut().push(msg.to_string());
}
fn get_messages(&self) -> Vec<String> {
self.messages.borrow().clone()
}
}
```
## Pin and Self-Referential Types
```rust
use std::pin::Pin;
use std::marker::PhantomPinned;
// Self-referential struct (requires Pin)
struct SelfReferential {
data: String,
pointer: *const String,
_pin: PhantomPinned,
}
impl SelfReferential {
fn new(data: String) -> Pin<Box<Self>> {
let mut boxed = Box::pin(Self {
data,
pointer: std::ptr::null(),
_pin: PhantomPinned,
});
// Safe: we're not moving the data after this
let ptr = &boxed.data as *const String;
unsafe {
let mut_ref = Pin::as_mut(&mut boxed);
Pin::get_unchecked_mut(mut_ref).pointer = ptr;
}
boxed
}
}
// Pin in async contexts
async fn pinned_future() {
// Futures are often self-referential, hence Pin
let fut = async { 42 };
let pinned = Box::pin(fut);
pinned.await;
}
```
## Cow (Clone on Write)
```rust
use std::borrow::Cow;
fn process_text(input: &str) -> Cow<str> {
if input.contains("bad") {
// Need to modify: allocate new String
Cow::Owned(input.replace("bad", "good"))
} else {
// No modification needed: just borrow
Cow::Borrowed(input)
}
}
// Usage
let text1 = "hello world";
let result1 = process_text(text1); // Borrowed (no allocation)
let text2 = "bad word";
let result2 = process_text(text2); // Owned (allocated)
```
## Drop Trait and RAII
```rust
struct FileGuard {
name: String,
}
impl FileGuard {
fn new(name: String) -> Self {
println!("Opening {}", name);
Self { name }
}
}
impl Drop for FileGuard {
fn drop(&mut self) {
println!("Closing {}", self.name);
}
}
// Usage: automatic cleanup
{
let _file = FileGuard::new("data.txt".to_string());
// Use file...
} // Drop called automatically here
```
## Common Patterns
```rust
// Builder pattern with ownership
struct Config {
host: String,
port: u16,
}
impl Config {
fn builder() -> ConfigBuilder {
ConfigBuilder::default()
}
}
struct ConfigBuilder {
host: Option<String>,
port: Option<u16>,
}
impl ConfigBuilder {
fn host(mut self, host: impl Into<String>) -> Self {
self.host = Some(host.into());
self
}
fn port(mut self, port: u16) -> Self {
self.port = Some(port);
self
}
fn build(self) -> Result<Config, &'static str> {
Ok(Config {
host: self.host.ok_or("host required")?,
port: self.port.unwrap_or(8080),
})
}
}
// Usage
let config = Config::builder()
.host("localhost")
.port(3000)
.build()?;
```
## Best Practices
- Prefer borrowing (&T) over ownership transfer when possible
- Use &str over String for function parameters
- Use &[T] over Vec<T> for function parameters
- Clone only when necessary (profile first)
- Use Cow<'a, T> for conditional cloning
- Document lifetime relationships in complex cases
- Use Arc<Mutex<T>> for shared mutable state across threads
- Use Rc<RefCell<T>> for shared mutable state in single thread
- Implement Drop for RAII patterns
- Use PhantomData to constrain variance when needed

470
skills/rust-engineer/references/testing.md
View File

@ -1,470 +0,0 @@
# Testing in Rust
## Unit Tests
```rust
// Tests in same file
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_addition() {
assert_eq!(2 + 2, 4);
}
#[test]
fn test_subtraction() {
assert!(10 - 5 == 5);
}
#[test]
#[should_panic(expected = "division by zero")]
fn test_panic() {
divide(10, 0);
}
#[test]
fn test_result() -> Result<(), String> {
let result = divide(10, 2)?;
assert_eq!(result, 5);
Ok(())
}
#[test]
#[ignore]
fn expensive_test() {
// Run with: cargo test -- --ignored
}
}
// Assertions
fn assert_examples() {
assert!(true);
assert_eq!(2 + 2, 4);
assert_ne!(2 + 2, 5);
// Custom messages
assert!(value > 0, "Value must be positive, got {}", value);
assert_eq!(result, expected, "Calculation failed");
}
```
## Doctests
```rust
/// Adds two numbers together.
///
/// # Examples
///
/// ```
/// use mylib::add;
///
/// let result = add(2, 3);
/// assert_eq!(result, 5);
/// ```
///
/// ```should_panic
/// use mylib::divide;
///
/// divide(10, 0); // This will panic
/// ```
///
/// ```ignore
/// // This code won't compile but won't fail the test
/// let x = undefined_function();
/// ```
pub fn add(a: i32, b: i32) -> i32 {
a + b
}
```
## Integration Tests
```rust
// tests/integration_test.rs
use mylib;
#[test]
fn test_full_workflow() {
let config = mylib::Config::new("test.conf");
let result = mylib::process(&config);
assert!(result.is_ok());
}
// tests/common/mod.rs - shared test utilities
pub fn setup() -> TestContext {
TestContext {
db: create_test_db(),
}
}
// tests/another_test.rs
mod common;
#[test]
fn test_with_common() {
let ctx = common::setup();
// Use ctx...
}
```
## Test Organization
```rust
// Nested test modules
#[cfg(test)]
mod tests {
use super::*;
mod addition {
use super::*;
#[test]
fn positive_numbers() {
assert_eq!(add(2, 3), 5);
}
#[test]
fn negative_numbers() {
assert_eq!(add(-2, -3), -5);
}
}
mod subtraction {
use super::*;
#[test]
fn test_subtract() {
assert_eq!(subtract(10, 5), 5);
}
}
}
```
## Test Fixtures and Setup
```rust
struct TestContext {
temp_dir: std::path::PathBuf,
db: Database,
}
impl TestContext {
fn setup() -> Self {
let temp_dir = std::env::temp_dir().join("test");
std::fs::create_dir_all(&temp_dir).unwrap();
Self {
temp_dir,
db: Database::connect_test(),
}
}
}
impl Drop for TestContext {
fn drop(&mut self) {
// Cleanup
std::fs::remove_dir_all(&self.temp_dir).ok();
self.db.disconnect();
}
}
#[test]
fn test_with_fixture() {
let ctx = TestContext::setup();
// Test uses ctx...
// Automatic cleanup via Drop
}
```
## Async Tests
```rust
use tokio;
#[tokio::test]
async fn test_async_function() {
let result = async_operation().await;
assert_eq!(result, 42);
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn test_with_custom_runtime() {
let result = concurrent_operation().await;
assert!(result.is_ok());
}
// Testing async with timeout
#[tokio::test]
async fn test_with_timeout() {
let timeout = std::time::Duration::from_secs(5);
let result = tokio::time::timeout(timeout, slow_operation()).await;
assert!(result.is_ok());
}
```
## Property-Based Testing (proptest)
```rust
use proptest::prelude::*;
// Simple property test
proptest! {
#[test]
fn test_reversing_twice_is_identity(ref s in ".*") {
let reversed: String = s.chars().rev().collect();
let double_reversed: String = reversed.chars().rev().collect();
assert_eq!(s, &double_reversed);
}
}
// Custom strategies
proptest! {
#[test]
fn test_addition_commutative(a in 0..1000i32, b in 0..1000i32) {
assert_eq!(a + b, b + a);
}
#[test]
fn test_vector_push_pop(
ref v in prop::collection::vec(0..100i32, 0..100),
item in 0..100i32
) {
let mut v = v.clone();
v.push(item);
assert_eq!(v.pop(), Some(item));
}
}
// Complex custom strategies
fn user_strategy() -> impl Strategy<Value = User> {
(1..1000u64, "[a-z]{3,10}", "[a-z0-9.]+@[a-z]+\\.[a-z]+")
.prop_map(|(id, name, email)| User { id, name, email })
}
proptest! {
#[test]
fn test_user_serialization(user in user_strategy()) {
let json = serde_json::to_string(&user).unwrap();
let deserialized: User = serde_json::from_str(&json).unwrap();
assert_eq!(user, deserialized);
}
}
```
## Mocking
```rust
// Using mockall
use mockall::*;
use mockall::predicate::*;
#[automock]
trait Database {
fn get_user(&self, id: u64) -> Option<User>;
fn save_user(&mut self, user: User) -> Result<(), Error>;
}
#[test]
fn test_with_mock() {
let mut mock = MockDatabase::new();
mock.expect_get_user()
.with(eq(1))
.times(1)
.returning(|_| Some(User { id: 1, name: "Alice".to_string() }));
mock.expect_save_user()
.times(1)
.returning(|_| Ok(()));
// Use mock in test
let user = mock.get_user(1);
assert!(user.is_some());
}
```
## Benchmarks (Criterion)
```rust
// benches/my_benchmark.rs
use criterion::{black_box, criterion_group, criterion_main, Criterion};
fn fibonacci(n: u64) -> u64 {
match n {
0 => 1,
1 => 1,
n => fibonacci(n - 1) + fibonacci(n - 2),
}
}
fn criterion_benchmark(c: &mut Criterion) {
c.bench_function("fib 20", |b| b.iter(|| fibonacci(black_box(20))));
}
criterion_group!(benches, criterion_benchmark);
criterion_main!(benches);
// Cargo.toml:
// [dev-dependencies]
// criterion = "0.5"
//
// [[bench]]
// name = "my_benchmark"
// harness = false
```
## Advanced Benchmarking
```rust
use criterion::{BenchmarkId, Criterion, criterion_group, criterion_main};
fn bench_multiple_sizes(c: &mut Criterion) {
let mut group = c.benchmark_group("sorting");
for size in [10, 100, 1000, 10000].iter() {
group.bench_with_input(BenchmarkId::from_parameter(size), size, |b, &size| {
b.iter_batched(
|| generate_random_vec(size),
|mut v| v.sort(),
criterion::BatchSize::SmallInput,
);
});
}
group.finish();
}
// Comparing implementations
fn bench_comparison(c: &mut Criterion) {
let mut group = c.benchmark_group("string_search");
group.bench_function("naive", |b| {
b.iter(|| naive_search(black_box("haystack"), black_box("needle")))
});
group.bench_function("optimized", |b| {
b.iter(|| optimized_search(black_box("haystack"), black_box("needle")))
});
group.finish();
}
criterion_group!(benches, bench_multiple_sizes, bench_comparison);
criterion_main!(benches);
```
## Testing with External Resources
```rust
// Testing file I/O
#[test]
fn test_file_operations() {
use std::io::Write;
let temp_dir = std::env::temp_dir();
let file_path = temp_dir.join("test_file.txt");
// Write
let mut file = std::fs::File::create(&file_path).unwrap();
file.write_all(b"test content").unwrap();
// Read
let content = std::fs::read_to_string(&file_path).unwrap();
assert_eq!(content, "test content");
// Cleanup
std::fs::remove_file(&file_path).unwrap();
}
// Testing with databases (using sqlx)
#[sqlx::test]
async fn test_database_operations(pool: sqlx::PgPool) -> sqlx::Result<()> {
sqlx::query("INSERT INTO users (name) VALUES ($1)")
.bind("Alice")
.execute(&pool)
.await?;
let count: (i64,) = sqlx::query_as("SELECT COUNT(*) FROM users")
.fetch_one(&pool)
.await?;
assert_eq!(count.0, 1);
Ok(())
}
```
## Snapshot Testing
```rust
// Using insta crate
use insta::assert_snapshot;
#[test]
fn test_output_format() {
let data = generate_complex_output();
assert_snapshot!(data);
}
#[test]
fn test_json_output() {
let json = serde_json::to_string_pretty(&get_data()).unwrap();
assert_snapshot!(json);
}
// Run with: cargo insta test
// Review snapshots: cargo insta review
```
## Code Coverage
```rust
// Using tarpaulin
// cargo install cargo-tarpaulin
// cargo tarpaulin --out Html --output-dir coverage
// Using llvm-cov
// cargo install cargo-llvm-cov
// cargo llvm-cov --html
```
## Fuzzing
```rust
// Using cargo-fuzz
// cargo install cargo-fuzz
// cargo fuzz init
// fuzz/fuzz_targets/fuzz_target_1.rs
#![no_main]
use libfuzzer_sys::fuzz_target;
fuzz_target!(|data: &[u8]| {
if let Ok(s) = std::str::from_utf8(data) {
let _ = mylib::parse_input(s);
}
});
// Run with: cargo fuzz run fuzz_target_1
```
## Best Practices
- Write tests alongside production code in #[cfg(test)] modules
- Use integration tests in tests/ directory for end-to-end testing
- Include doctests in documentation for examples that must work
- Use descriptive test names that explain what is being tested
- Test edge cases (empty inputs, max values, etc.)
- Use property-based testing for algorithmic code
- Benchmark performance-critical code with criterion
- Run tests in CI with cargo test --all-features
- Use cargo test -- --nocapture to see println! output
- Test error conditions with #[should_panic] or Result
- Mock external dependencies for unit tests
- Use test fixtures for complex setup/teardown
- Run clippy on test code too
- Measure code coverage and aim for high coverage
- Use fuzzing for security-critical parsers
- Test async code with tokio::test
- Use snapshot testing for complex output validation

413
skills/rust-engineer/references/traits.md
View File

@ -1,413 +0,0 @@
# Traits, Generics, and Type System
## Basic Trait Definition
```rust
// Simple trait
trait Drawable {
fn draw(&self);
}
// Trait with default implementation
trait Describable {
fn describe(&self) -> String {
String::from("No description available")
}
}
// Implementing traits
struct Circle {
radius: f64,
}
impl Drawable for Circle {
fn draw(&self) {
println!("Drawing circle with radius {}", self.radius);
}
}
impl Describable for Circle {
fn describe(&self) -> String {
format!("A circle with radius {}", self.radius)
}
}
```
## Associated Types
```rust
// Associated types vs generic parameters
trait Container {
type Item;
fn add(&mut self, item: Self::Item);
fn get(&self, index: usize) -> Option<&Self::Item>;
}
impl Container for Vec<i32> {
type Item = i32;
fn add(&mut self, item: i32) {
self.push(item);
}
fn get(&self, index: usize) -> Option<&i32> {
self.get(index)
}
}
// Iterator trait (standard library example)
trait MyIterator {
type Item;
fn next(&mut self) -> Option<Self::Item>;
}
```
## Generic Traits and Bounds
```rust
// Generic trait with multiple bounds
fn print_info<T>(item: &T)
where
T: std::fmt::Display + std::fmt::Debug,
{
println!("Display: {}", item);
println!("Debug: {:?}", item);
}
// Generic struct with trait bounds
struct Pair<T: PartialOrd> {
first: T,
second: T,
}
impl<T: PartialOrd> Pair<T> {
fn new(first: T, second: T) -> Self {
Self { first, second }
}
fn larger(&self) -> &T {
if self.first > self.second {
&self.first
} else {
&self.second
}
}
}
// Blanket implementation
trait MyTrait {
fn do_something(&self);
}
impl<T: std::fmt::Display> MyTrait for T {
fn do_something(&self) {
println!("Value: {}", self);
}
}
```
## Trait Objects (Dynamic Dispatch)
```rust
// Static dispatch (monomorphization)
fn static_dispatch<T: Drawable>(item: &T) {
item.draw();
}
// Dynamic dispatch (trait objects)
fn dynamic_dispatch(item: &dyn Drawable) {
item.draw();
}
// Storing trait objects
struct Canvas {
shapes: Vec<Box<dyn Drawable>>,
}
impl Canvas {
fn new() -> Self {
Self { shapes: Vec::new() }
}
fn add_shape(&mut self, shape: Box<dyn Drawable>) {
self.shapes.push(shape);
}
fn draw_all(&self) {
for shape in &self.shapes {
shape.draw();
}
}
}
// Object safety: traits must meet criteria
trait ObjectSafe {
fn method(&self); // OK: takes &self
}
trait NotObjectSafe {
fn generic<T>(&self); // NOT OK: generic method
fn by_value(self); // NOT OK: takes self by value
}
```
## Derive Macros
```rust
// Standard derive macros
#[derive(Debug, Clone, PartialEq, Eq, Hash)]
struct User {
id: u64,
name: String,
}
// Deriving more traits
#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord)]
struct Point {
x: i32,
y: i32,
}
// Custom derive with serde
use serde::{Deserialize, Serialize};
#[derive(Debug, Serialize, Deserialize)]
struct Config {
host: String,
port: u16,
}
```
## Advanced Trait Patterns
```rust
// Extension trait pattern
trait StringExt {
fn truncate_to(&self, max_len: usize) -> String;
}
impl StringExt for str {
fn truncate_to(&self, max_len: usize) -> String {
if self.len() <= max_len {
self.to_string()
} else {
format!("{}...", &self[..max_len])
}
}
}
// Sealed trait pattern (prevent external implementation)
mod sealed {
pub trait Sealed {}
}
pub trait MySealed: sealed::Sealed {
fn method(&self);
}
struct MyType;
impl sealed::Sealed for MyType {}
impl MySealed for MyType {
fn method(&self) {
println!("Implemented");
}
}
// Supertraits
trait Printable {
fn print(&self);
}
trait Loggable: Printable { // Supertrait: must also impl Printable
fn log(&self) {
self.print(); // Can call supertrait methods
}
}
```
## Associated Constants
```rust
trait Config {
const MAX_SIZE: usize;
const DEFAULT_TIMEOUT: u64;
}
struct ServerConfig;
impl Config for ServerConfig {
const MAX_SIZE: usize = 1024;
const DEFAULT_TIMEOUT: u64 = 30;
}
fn use_config<T: Config>() {
println!("Max size: {}", T::MAX_SIZE);
}
```
## Generic Associated Types (GATs)
```rust
// GATs allow generics in associated types
trait LendingIterator {
type Item<'a> where Self: 'a;
fn next<'a>(&'a mut self) -> Option<Self::Item<'a>>;
}
struct WindowsMut<'data, T> {
data: &'data mut [T],
index: usize,
}
impl<'data, T> LendingIterator for WindowsMut<'data, T> {
type Item<'a> = &'a mut [T] where Self: 'a;
fn next<'a>(&'a mut self) -> Option<Self::Item<'a>> {
if self.index >= self.data.len() {
return None;
}
let start = self.index;
self.index += 2;
Some(&mut self.data[start..start.min(self.data.len())])
}
}
```
## Marker Traits
```rust
use std::marker::{PhantomData, Send, Sync};
// Send: type can be transferred across thread boundaries
// Sync: type can be shared between threads (&T is Send)
// Custom marker trait
trait Trusted {}
struct TrustedData<T> {
data: T,
_marker: PhantomData<T>,
}
impl<T: Trusted> TrustedData<T> {
fn new(data: T) -> Self {
Self {
data,
_marker: PhantomData,
}
}
}
```
## Operator Overloading
```rust
use std::ops::{Add, Mul};
#[derive(Debug, Clone, Copy)]
struct Vector2D {
x: f64,
y: f64,
}
impl Add for Vector2D {
type Output = Self;
fn add(self, other: Self) -> Self {
Self {
x: self.x + other.x,
y: self.y + other.y,
}
}
}
impl Mul<f64> for Vector2D {
type Output = Self;
fn mul(self, scalar: f64) -> Self {
Self {
x: self.x * scalar,
y: self.y * scalar,
}
}
}
// Usage
let v1 = Vector2D { x: 1.0, y: 2.0 };
let v2 = Vector2D { x: 3.0, y: 4.0 };
let v3 = v1 + v2;
let v4 = v1 * 2.5;
```
## From/Into Conversion Traits
```rust
struct UserId(u64);
impl From<u64> for UserId {
fn from(id: u64) -> Self {
UserId(id)
}
}
// Into is automatically implemented
fn accept_user_id(id: impl Into<UserId>) {
let user_id = id.into();
println!("User ID: {}", user_id.0);
}
// TryFrom for fallible conversions
use std::convert::TryFrom;
impl TryFrom<i64> for UserId {
type Error = &'static str;
fn try_from(value: i64) -> Result<Self, Self::Error> {
if value < 0 {
Err("User ID cannot be negative")
} else {
Ok(UserId(value as u64))
}
}
}
```
## Const Traits (Nightly)
```rust
// Const trait implementations (requires nightly)
#![feature(const_trait_impl)]
#[const_trait]
trait ConstAdd {
fn add(self, other: Self) -> Self;
}
impl const ConstAdd for i32 {
fn add(self, other: Self) -> Self {
self + other
}
}
const fn compute() -> i32 {
5.add(10) // Can use in const context
}
```
## Best Practices
- Prefer associated types when there's one clear type per implementation
- Use generic parameters when multiple types might be used simultaneously
- Keep traits small and focused (single responsibility)
- Use extension traits to add functionality to existing types
- Document trait requirements and invariants
- Use marker traits for compile-time guarantees
- Prefer static dispatch for performance, dynamic dispatch for flexibility
- Use #[derive] when possible instead of manual implementations
- Implement standard traits (Debug, Clone, etc.) for better ecosystem integration
- Use sealed traits to prevent external implementations when needed

252
skills/setup-browser-cookies/SKILL.md
View File

@ -1,252 +0,0 @@
---
name: setup-browser-cookies
version: 1.0.0
description: |
Import cookies from your real browser (Comet, Chrome, Arc, Brave, Edge) into the
headless browse session. Opens an interactive picker UI where you select which
cookie domains to import. Use before QA testing authenticated pages. Use when asked
to "import cookies", "login to the site", or "authenticate the browser".
maturity: imported
allowed-tools:
- Bash
- Read
- AskUserQuestion
---
<!-- AUTO-GENERATED from SKILL.md.tmpl — do not edit directly -->
<!-- Regenerate: bun run gen:skill-docs -->
## Preamble (run first)
```bash
_UPD=$(~/.claude/skills/gstack/bin/gstack-update-check 2>/dev/null || .claude/skills/gstack/bin/gstack-update-check 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
mkdir -p ~/.gstack/sessions
touch ~/.gstack/sessions/"$PPID"
_SESSIONS=$(find ~/.gstack/sessions -mmin -120 -type f 2>/dev/null | wc -l | tr -d ' ')
find ~/.gstack/sessions -mmin +120 -type f -delete 2>/dev/null || true
_CONTRIB=$(~/.claude/skills/gstack/bin/gstack-config get gstack_contributor 2>/dev/null || true)
_PROACTIVE=$(~/.claude/skills/gstack/bin/gstack-config get proactive 2>/dev/null || echo "true")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
echo "BRANCH: $_BRANCH"
echo "PROACTIVE: $_PROACTIVE"
_LAKE_SEEN=$([ -f ~/.gstack/.completeness-intro-seen ] && echo "yes" || echo "no")
echo "LAKE_INTRO: $_LAKE_SEEN"
```
If `PROACTIVE` is `"false"`, do not proactively suggest gstack skills — only invoke
them when the user explicitly asks. The user opted out of proactive suggestions.
If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue.
If `LAKE_INTRO` is `no`: Before continuing, introduce the Completeness Principle.
Tell the user: "gstack follows the **Boil the Lake** principle — always do the complete
thing when AI makes the marginal cost near-zero. Read more: https://garryslist.org/posts/boil-the-ocean"
Then offer to open the essay in their default browser:
```bash
open https://garryslist.org/posts/boil-the-ocean
touch ~/.gstack/.completeness-intro-seen
```
Only run `open` if the user says yes. Always run `touch` to mark as seen. This only happens once.
## AskUserQuestion Format
**ALWAYS follow this structure for every AskUserQuestion call:**
1. **Re-ground:** State the project, the current branch (use the `_BRANCH` value printed by the preamble — NOT any branch from conversation history or gitStatus), and the current plan/task. (1-2 sentences)
2. **Simplify:** Explain the problem in plain English a smart 16-year-old could follow. No raw function names, no internal jargon, no implementation details. Use concrete examples and analogies. Say what it DOES, not what it's called.
3. **Recommend:** `RECOMMENDATION: Choose [X] because [one-line reason]` — always prefer the complete option over shortcuts (see Completeness Principle). Include `Completeness: X/10` for each option. Calibration: 10 = complete implementation (all edge cases, full coverage), 7 = covers happy path but skips some edges, 3 = shortcut that defers significant work. If both options are 8+, pick the higher; if one is ≤5, flag it.
4. **Options:** Lettered options: `A) ... B) ... C) ...` — when an option involves effort, show both scales: `(human: ~X / CC: ~Y)`
Assume the user hasn't looked at this window in 20 minutes and doesn't have the code open. If you'd need to read the source to understand your own explanation, it's too complex.
Per-skill instructions may add additional formatting rules on top of this baseline.
## Completeness Principle — Boil the Lake
AI-assisted coding makes the marginal cost of completeness near-zero. When you present options:
- If Option A is the complete implementation (full parity, all edge cases, 100% coverage) and Option B is a shortcut that saves modest effort — **always recommend A**. The delta between 80 lines and 150 lines is meaningless with CC+gstack. "Good enough" is the wrong instinct when "complete" costs minutes more.
- **Lake vs. ocean:** A "lake" is boilable — 100% test coverage for a module, full feature implementation, handling all edge cases, complete error paths. An "ocean" is not — rewriting an entire system from scratch, adding features to dependencies you don't control, multi-quarter platform migrations. Recommend boiling lakes. Flag oceans as out of scope.
- **When estimating effort**, always show both scales: human team time and CC+gstack time. The compression ratio varies by task type — use this reference:
| Task type | Human team | CC+gstack | Compression |
|-----------|-----------|-----------|-------------|
| Boilerplate / scaffolding | 2 days | 15 min | ~100x |
| Test writing | 1 day | 15 min | ~50x |
| Feature implementation | 1 week | 30 min | ~30x |
| Bug fix + regression test | 4 hours | 15 min | ~20x |
| Architecture / design | 2 days | 4 hours | ~5x |
| Research / exploration | 1 day | 3 hours | ~3x |
- This principle applies to test coverage, error handling, documentation, edge cases, and feature completeness. Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
**Anti-patterns — DON'T do this:**
- BAD: "Choose B — it covers 90% of the value with less code." (If A is only 70 lines more, choose A.)
- BAD: "We can skip edge case handling to save time." (Edge case handling costs minutes with CC.)
- BAD: "Let's defer test coverage to a follow-up PR." (Tests are the cheapest lake to boil.)
- BAD: Quoting only human-team effort: "This would take 2 weeks." (Say: "2 weeks human / ~1 hour CC.")
## Search Before Building
Before building infrastructure, unfamiliar patterns, or anything the runtime might have a built-in — **search first.** Read `~/.claude/skills/gstack/ETHOS.md` for the full philosophy.
**Three layers of knowledge:**
- **Layer 1** (tried and true — in distribution). Don't reinvent the wheel. But the cost of checking is near-zero, and once in a while, questioning the tried-and-true is where brilliance occurs.
- **Layer 2** (new and popular — search for these). But scrutinize: humans are subject to mania. Search results are inputs to your thinking, not answers.
- **Layer 3** (first principles — prize these above all). Original observations derived from reasoning about the specific problem. The most valuable of all.
**Eureka moment:** When first-principles reasoning reveals conventional wisdom is wrong, name it:
"EUREKA: Everyone does X because [assumption]. But [evidence] shows this is wrong. Y is better because [reasoning]."
Log eureka moments:
```bash
jq -n --arg ts "$(date -u +%Y-%m-%dT%H:%M:%SZ)" --arg skill "SKILL_NAME" --arg branch "$(git branch --show-current 2>/dev/null)" --arg insight "ONE_LINE_SUMMARY" '{ts:$ts,skill:$skill,branch:$branch,insight:$insight}' >> ~/.gstack/analytics/eureka.jsonl 2>/dev/null || true
```
Replace SKILL_NAME and ONE_LINE_SUMMARY. Runs inline — don't stop the workflow.
**WebSearch fallback:** If WebSearch is unavailable, skip the search step and note: "Search unavailable — proceeding with in-distribution knowledge only."
## Contributor Mode
If `_CONTRIB` is `true`: you are in **contributor mode**. You're a gstack user who also helps make it better.
**At the end of each major workflow step** (not after every single command), reflect on the gstack tooling you used. Rate your experience 0 to 10. If it wasn't a 10, think about why. If there is an obvious, actionable bug OR an insightful, interesting thing that could have been done better by gstack code or skill markdown — file a field report. Maybe our contributor will help make us better!
**Calibration — this is the bar:** For example, `$B js "await fetch(...)"` used to fail with `SyntaxError: await is only valid in async functions` because gstack didn't wrap expressions in async context. Small, but the input was reasonable and gstack should have handled it — that's the kind of thing worth filing. Things less consequential than this, ignore.
**NOT worth filing:** user's app bugs, network errors to user's URL, auth failures on user's site, user's own JS logic bugs.
**To file:** write `~/.gstack/contributor-logs/{slug}.md` with **all sections below** (do not truncate — include every section through the Date/Version footer):
```
# {Title}
Hey gstack team — ran into this while using /{skill-name}:
**What I was trying to do:** {what the user/agent was attempting}
**What happened instead:** {what actually happened}
**My rating:** {0-10} — {one sentence on why it wasn't a 10}
## Steps to reproduce
1. {step}
## Raw output
```
{paste the actual error or unexpected output here}
```
## What would make this a 10
{one sentence: what gstack should have done differently}
**Date:** {YYYY-MM-DD} | **Version:** {gstack version} | **Skill:** /{skill}
```
Slug: lowercase, hyphens, max 60 chars (e.g. `browse-js-no-await`). Skip if file already exists. Max 3 reports per session. File inline and continue — don't stop the workflow. Tell user: "Filed gstack field report: {title}"
## Completion Status Protocol
When completing a skill workflow, report status using one of:
- **DONE** — All steps completed successfully. Evidence provided for each claim.
- **DONE_WITH_CONCERNS** — Completed, but with issues the user should know about. List each concern.
- **BLOCKED** — Cannot proceed. State what is blocking and what was tried.
- **NEEDS_CONTEXT** — Missing information required to continue. State exactly what you need.
### Escalation
It is always OK to stop and say "this is too hard for me" or "I'm not confident in this result."
Bad work is worse than no work. You will not be penalized for escalating.
- If you have attempted a task 3 times without success, STOP and escalate.
- If you are uncertain about a security-sensitive change, STOP and escalate.
- If the scope of work exceeds what you can verify, STOP and escalate.
Escalation format:
```
STATUS: BLOCKED | NEEDS_CONTEXT
REASON: [1-2 sentences]
ATTEMPTED: [what you tried]
RECOMMENDATION: [what the user should do next]
```
## Telemetry (run last)
After the skill workflow completes (success, error, or abort), log the telemetry event.
Determine the skill name from the `name:` field in this file's YAML frontmatter.
# Setup Browser Cookies
Import logged-in sessions from your real Chromium browser into the headless browse session.
## How it works
1. Find the browse binary
2. Run `cookie-import-browser` to detect installed browsers and open the picker UI
3. User selects which cookie domains to import in their browser
4. Cookies are decrypted and loaded into the Playwright session
## Steps
### 1. Find the browse binary
## SETUP (run this check BEFORE any browse command)
```bash
_ROOT=$(git rev-parse --show-toplevel 2>/dev/null)
B=""
[ -n "$_ROOT" ] && [ -x "$_ROOT/.claude/skills/gstack/browse/dist/browse" ] && B="$_ROOT/.claude/skills/gstack/browse/dist/browse"
[ -z "$B" ] && B=~/.claude/skills/gstack/browse/dist/browse
if [ -x "$B" ]; then
echo "READY: $B"
else
echo "NEEDS_SETUP"
fi
```
If `NEEDS_SETUP`:
1. Tell the user: "gstack browse needs a one-time build (~10 seconds). OK to proceed?" Then STOP and wait.
2. Run: `cd <SKILL_DIR> && ./setup`
3. If `bun` is not installed: `curl -fsSL https://bun.sh/install | bash`
### 2. Open the cookie picker
```bash
$B cookie-import-browser
```
This auto-detects installed Chromium browsers (Comet, Chrome, Arc, Brave, Edge) and opens
an interactive picker UI in your default browser where you can:
- Switch between installed browsers
- Search domains
- Click "+" to import a domain's cookies
- Click trash to remove imported cookies
Tell the user: **"Cookie picker opened — select the domains you want to import in your browser, then tell me when you're done."**
### 3. Direct import (alternative)
If the user specifies a domain directly (e.g., `/setup-browser-cookies github.com`), skip the UI:
```bash
$B cookie-import-browser comet --domain github.com
```
Replace `comet` with the appropriate browser if specified.
### 4. Verify
After the user confirms they're done:
```bash
$B cookies
```
Show the user a summary of imported cookies (domain counts).
## Notes
- First import per browser may trigger a macOS Keychain dialog — click "Allow" / "Always Allow"
- Cookie picker is served on the same port as the browse server (no extra process)
- Only domain names and cookie counts are shown in the UI — no cookie values are exposed
- The browse session persists cookies between commands, so imported cookies work immediately

74
skills/setup-browser-cookies/SKILL.md.tmpl
View File

@ -1,74 +0,0 @@
---
name: setup-browser-cookies
version: 1.0.0
description: |
Import cookies from your real browser (Comet, Chrome, Arc, Brave, Edge) into the
headless browse session. Opens an interactive picker UI where you select which
cookie domains to import. Use before QA testing authenticated pages. Use when asked
to "import cookies", "login to the site", or "authenticate the browser".
allowed-tools:
- Bash
- Read
- AskUserQuestion
---
{{PREAMBLE}}
# Setup Browser Cookies
Import logged-in sessions from your real Chromium browser into the headless browse session.
## How it works
1. Find the browse binary
2. Run `cookie-import-browser` to detect installed browsers and open the picker UI
3. User selects which cookie domains to import in their browser
4. Cookies are decrypted and loaded into the Playwright session
## Steps
### 1. Find the browse binary
{{BROWSE_SETUP}}
### 2. Open the cookie picker
```bash
$B cookie-import-browser
```
This auto-detects installed Chromium browsers (Comet, Chrome, Arc, Brave, Edge) and opens
an interactive picker UI in your default browser where you can:
- Switch between installed browsers
- Search domains
- Click "+" to import a domain's cookies
- Click trash to remove imported cookies
Tell the user: **"Cookie picker opened — select the domains you want to import in your browser, then tell me when you're done."**
### 3. Direct import (alternative)
If the user specifies a domain directly (e.g., `/setup-browser-cookies github.com`), skip the UI:
```bash
$B cookie-import-browser comet --domain github.com
```
Replace `comet` with the appropriate browser if specified.
### 4. Verify
After the user confirms they're done:
```bash
$B cookies
```
Show the user a summary of imported cookies (domain counts).
## Notes
- First import per browser may trigger a macOS Keychain dialog — click "Allow" / "Always Allow"
- Cookie picker is served on the same port as the browse server (no extra process)
- Only domain names and cookie counts are shown in the UI — no cookie values are exposed
- The browse session persists cookies between commands, so imported cookies work immediately

381
skills/setup-deploy/SKILL.md
View File

@ -1,381 +0,0 @@
---
name: setup-deploy
version: 1.0.0
description: |
Configure deployment settings for /land-and-deploy. Detects your deploy
platform (Fly.io, Render, Vercel, Netlify, Heroku, GitHub Actions, custom),
production URL, health check endpoints, and deploy status commands. Writes
the configuration to CLAUDE.md so all future deploys are automatic.
Use when: "setup deploy", "configure deployment", "set up land-and-deploy",
"how do I deploy with gstack", "add deploy config".
maturity: imported
allowed-tools:
- Bash
- Read
- Write
- Edit
- Glob
- Grep
- AskUserQuestion
---
<!-- AUTO-GENERATED from SKILL.md.tmpl — do not edit directly -->
<!-- Regenerate: bun run gen:skill-docs -->
## Preamble (run first)
```bash
_UPD=$(~/.claude/skills/gstack/bin/gstack-update-check 2>/dev/null || .claude/skills/gstack/bin/gstack-update-check 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
mkdir -p ~/.gstack/sessions
touch ~/.gstack/sessions/"$PPID"
_SESSIONS=$(find ~/.gstack/sessions -mmin -120 -type f 2>/dev/null | wc -l | tr -d ' ')
find ~/.gstack/sessions -mmin +120 -type f -delete 2>/dev/null || true
_CONTRIB=$(~/.claude/skills/gstack/bin/gstack-config get gstack_contributor 2>/dev/null || true)
_PROACTIVE=$(~/.claude/skills/gstack/bin/gstack-config get proactive 2>/dev/null || echo "true")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
echo "BRANCH: $_BRANCH"
echo "PROACTIVE: $_PROACTIVE"
_LAKE_SEEN=$([ -f ~/.gstack/.completeness-intro-seen ] && echo "yes" || echo "no")
echo "LAKE_INTRO: $_LAKE_SEEN"
```
If `PROACTIVE` is `"false"`, do not proactively suggest gstack skills — only invoke
them when the user explicitly asks. The user opted out of proactive suggestions.
If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue.
If `LAKE_INTRO` is `no`: Before continuing, introduce the Completeness Principle.
Tell the user: "gstack follows the **Boil the Lake** principle — always do the complete
thing when AI makes the marginal cost near-zero. Read more: https://garryslist.org/posts/boil-the-ocean"
Then offer to open the essay in their default browser:
```bash
open https://garryslist.org/posts/boil-the-ocean
touch ~/.gstack/.completeness-intro-seen
```
Only run `open` if the user says yes. Always run `touch` to mark as seen. This only happens once.
## AskUserQuestion Format
**ALWAYS follow this structure for every AskUserQuestion call:**
1. **Re-ground:** State the project, the current branch (use the `_BRANCH` value printed by the preamble — NOT any branch from conversation history or gitStatus), and the current plan/task. (1-2 sentences)
2. **Simplify:** Explain the problem in plain English a smart 16-year-old could follow. No raw function names, no internal jargon, no implementation details. Use concrete examples and analogies. Say what it DOES, not what it's called.
3. **Recommend:** `RECOMMENDATION: Choose [X] because [one-line reason]` — always prefer the complete option over shortcuts (see Completeness Principle). Include `Completeness: X/10` for each option. Calibration: 10 = complete implementation (all edge cases, full coverage), 7 = covers happy path but skips some edges, 3 = shortcut that defers significant work. If both options are 8+, pick the higher; if one is ≤5, flag it.
4. **Options:** Lettered options: `A) ... B) ... C) ...` — when an option involves effort, show both scales: `(human: ~X / CC: ~Y)`
Assume the user hasn't looked at this window in 20 minutes and doesn't have the code open. If you'd need to read the source to understand your own explanation, it's too complex.
Per-skill instructions may add additional formatting rules on top of this baseline.
## Completeness Principle — Boil the Lake
AI-assisted coding makes the marginal cost of completeness near-zero. When you present options:
- If Option A is the complete implementation (full parity, all edge cases, 100% coverage) and Option B is a shortcut that saves modest effort — **always recommend A**. The delta between 80 lines and 150 lines is meaningless with CC+gstack. "Good enough" is the wrong instinct when "complete" costs minutes more.
- **Lake vs. ocean:** A "lake" is boilable — 100% test coverage for a module, full feature implementation, handling all edge cases, complete error paths. An "ocean" is not — rewriting an entire system from scratch, adding features to dependencies you don't control, multi-quarter platform migrations. Recommend boiling lakes. Flag oceans as out of scope.
- **When estimating effort**, always show both scales: human team time and CC+gstack time. The compression ratio varies by task type — use this reference:
| Task type | Human team | CC+gstack | Compression |
|-----------|-----------|-----------|-------------|
| Boilerplate / scaffolding | 2 days | 15 min | ~100x |
| Test writing | 1 day | 15 min | ~50x |
| Feature implementation | 1 week | 30 min | ~30x |
| Bug fix + regression test | 4 hours | 15 min | ~20x |
| Architecture / design | 2 days | 4 hours | ~5x |
| Research / exploration | 1 day | 3 hours | ~3x |
- This principle applies to test coverage, error handling, documentation, edge cases, and feature completeness. Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
**Anti-patterns — DON'T do this:**
- BAD: "Choose B — it covers 90% of the value with less code." (If A is only 70 lines more, choose A.)
- BAD: "We can skip edge case handling to save time." (Edge case handling costs minutes with CC.)
- BAD: "Let's defer test coverage to a follow-up PR." (Tests are the cheapest lake to boil.)
- BAD: Quoting only human-team effort: "This would take 2 weeks." (Say: "2 weeks human / ~1 hour CC.")
## Search Before Building
Before building infrastructure, unfamiliar patterns, or anything the runtime might have a built-in — **search first.** Read `~/.claude/skills/gstack/ETHOS.md` for the full philosophy.
**Three layers of knowledge:**
- **Layer 1** (tried and true — in distribution). Don't reinvent the wheel. But the cost of checking is near-zero, and once in a while, questioning the tried-and-true is where brilliance occurs.
- **Layer 2** (new and popular — search for these). But scrutinize: humans are subject to mania. Search results are inputs to your thinking, not answers.
- **Layer 3** (first principles — prize these above all). Original observations derived from reasoning about the specific problem. The most valuable of all.
**Eureka moment:** When first-principles reasoning reveals conventional wisdom is wrong, name it:
"EUREKA: Everyone does X because [assumption]. But [evidence] shows this is wrong. Y is better because [reasoning]."
Log eureka moments:
```bash
jq -n --arg ts "$(date -u +%Y-%m-%dT%H:%M:%SZ)" --arg skill "SKILL_NAME" --arg branch "$(git branch --show-current 2>/dev/null)" --arg insight "ONE_LINE_SUMMARY" '{ts:$ts,skill:$skill,branch:$branch,insight:$insight}' >> ~/.gstack/analytics/eureka.jsonl 2>/dev/null || true
```
Replace SKILL_NAME and ONE_LINE_SUMMARY. Runs inline — don't stop the workflow.
**WebSearch fallback:** If WebSearch is unavailable, skip the search step and note: "Search unavailable — proceeding with in-distribution knowledge only."
## Contributor Mode
If `_CONTRIB` is `true`: you are in **contributor mode**. You're a gstack user who also helps make it better.
**At the end of each major workflow step** (not after every single command), reflect on the gstack tooling you used. Rate your experience 0 to 10. If it wasn't a 10, think about why. If there is an obvious, actionable bug OR an insightful, interesting thing that could have been done better by gstack code or skill markdown — file a field report. Maybe our contributor will help make us better!
**Calibration — this is the bar:** For example, `$B js "await fetch(...)"` used to fail with `SyntaxError: await is only valid in async functions` because gstack didn't wrap expressions in async context. Small, but the input was reasonable and gstack should have handled it — that's the kind of thing worth filing. Things less consequential than this, ignore.
**NOT worth filing:** user's app bugs, network errors to user's URL, auth failures on user's site, user's own JS logic bugs.
**To file:** write `~/.gstack/contributor-logs/{slug}.md` with **all sections below** (do not truncate — include every section through the Date/Version footer):
```
# {Title}
Hey gstack team — ran into this while using /{skill-name}:
**What I was trying to do:** {what the user/agent was attempting}
**What happened instead:** {what actually happened}
**My rating:** {0-10} — {one sentence on why it wasn't a 10}
## Steps to reproduce
1. {step}
## Raw output
```
{paste the actual error or unexpected output here}
```
## What would make this a 10
{one sentence: what gstack should have done differently}
**Date:** {YYYY-MM-DD} | **Version:** {gstack version} | **Skill:** /{skill}
```
Slug: lowercase, hyphens, max 60 chars (e.g. `browse-js-no-await`). Skip if file already exists. Max 3 reports per session. File inline and continue — don't stop the workflow. Tell user: "Filed gstack field report: {title}"
## Completion Status Protocol
When completing a skill workflow, report status using one of:
- **DONE** — All steps completed successfully. Evidence provided for each claim.
- **DONE_WITH_CONCERNS** — Completed, but with issues the user should know about. List each concern.
- **BLOCKED** — Cannot proceed. State what is blocking and what was tried.
- **NEEDS_CONTEXT** — Missing information required to continue. State exactly what you need.
### Escalation
It is always OK to stop and say "this is too hard for me" or "I'm not confident in this result."
Bad work is worse than no work. You will not be penalized for escalating.
- If you have attempted a task 3 times without success, STOP and escalate.
- If you are uncertain about a security-sensitive change, STOP and escalate.
- If the scope of work exceeds what you can verify, STOP and escalate.
Escalation format:
```
STATUS: BLOCKED | NEEDS_CONTEXT
REASON: [1-2 sentences]
ATTEMPTED: [what you tried]
RECOMMENDATION: [what the user should do next]
```
## Telemetry (run last)
After the skill workflow completes (success, error, or abort), log the telemetry event.
Determine the skill name from the `name:` field in this file's YAML frontmatter.
# /setup-deploy — Configure Deployment for gstack
You are helping the user configure their deployment so `/land-and-deploy` works
automatically. Your job is to detect the deploy platform, production URL, health
checks, and deploy status commands — then persist everything to CLAUDE.md.
After this runs once, `/land-and-deploy` reads CLAUDE.md and skips detection entirely.
## User-invocable
When the user types `/setup-deploy`, run this skill.
## Instructions
### Step 1: Check existing configuration
```bash
grep -A 20 "## Deploy Configuration" CLAUDE.md 2>/dev/null || echo "NO_CONFIG"
```
If configuration already exists, show it and ask:
- **Context:** Deploy configuration already exists in CLAUDE.md.
- **RECOMMENDATION:** Choose A to update if your setup changed.
- A) Reconfigure from scratch (overwrite existing)
- B) Edit specific fields (show current config, let me change one thing)
- C) Done — configuration looks correct
If the user picks C, stop.
### Step 2: Detect platform
Run the platform detection from the deploy bootstrap:
```bash
# Platform config files
[ -f fly.toml ] && echo "PLATFORM:fly" && cat fly.toml
[ -f render.yaml ] && echo "PLATFORM:render" && cat render.yaml
[ -f vercel.json ] || [ -d .vercel ] && echo "PLATFORM:vercel"
[ -f netlify.toml ] && echo "PLATFORM:netlify" && cat netlify.toml
[ -f Procfile ] && echo "PLATFORM:heroku"
[ -f railway.json ] || [ -f railway.toml ] && echo "PLATFORM:railway"
# GitHub Actions deploy workflows
for f in .github/workflows/*.yml .github/workflows/*.yaml; do
[ -f "$f" ] && grep -qiE "deploy|release|production|staging|cd" "$f" 2>/dev/null && echo "DEPLOY_WORKFLOW:$f"
done
# Project type
[ -f package.json ] && grep -q '"bin"' package.json 2>/dev/null && echo "PROJECT_TYPE:cli"
ls *.gemspec 2>/dev/null && echo "PROJECT_TYPE:library"
```
### Step 3: Platform-specific setup
Based on what was detected, guide the user through platform-specific configuration.
#### Fly.io
If `fly.toml` detected:
1. Extract app name: `grep -m1 "^app" fly.toml | sed 's/app = "\(.*\)"/\1/'`
2. Check if `fly` CLI is installed: `which fly 2>/dev/null`
3. If installed, verify: `fly status --app {app} 2>/dev/null`
4. Infer URL: `https://{app}.fly.dev`
5. Set deploy status command: `fly status --app {app}`
6. Set health check: `https://{app}.fly.dev` (or `/health` if the app has one)
Ask the user to confirm the production URL. Some Fly apps use custom domains.
#### Render
If `render.yaml` detected:
1. Extract service name and type from render.yaml
2. Check for Render API key: `echo $RENDER_API_KEY | head -c 4` (don't expose the full key)
3. Infer URL: `https://{service-name}.onrender.com`
4. Render deploys automatically on push to the connected branch — no deploy workflow needed
5. Set health check: the inferred URL
Ask the user to confirm. Render uses auto-deploy from the connected git branch — after
merge to main, Render picks it up automatically. The "deploy wait" in /land-and-deploy
should poll the Render URL until it responds with the new version.
#### Vercel
If vercel.json or .vercel detected:
1. Check for `vercel` CLI: `which vercel 2>/dev/null`
2. If installed: `vercel ls --prod 2>/dev/null | head -3`
3. Vercel deploys automatically on push — preview on PR, production on merge to main
4. Set health check: the production URL from vercel project settings
#### Netlify
If netlify.toml detected:
1. Extract site info from netlify.toml
2. Netlify deploys automatically on push
3. Set health check: the production URL
#### GitHub Actions only
If deploy workflows detected but no platform config:
1. Read the workflow file to understand what it does
2. Extract the deploy target (if mentioned)
3. Ask the user for the production URL
#### Custom / Manual
If nothing detected:
Use AskUserQuestion to gather the information:
1. **How are deploys triggered?**
- A) Automatically on push to main (Fly, Render, Vercel, Netlify, etc.)
- B) Via GitHub Actions workflow
- C) Via a deploy script or CLI command (describe it)
- D) Manually (SSH, dashboard, etc.)
- E) This project doesn't deploy (library, CLI, tool)
2. **What's the production URL?** (Free text — the URL where the app runs)
3. **How can gstack check if a deploy succeeded?**
- A) HTTP health check at a specific URL (e.g., /health, /api/status)
- B) CLI command (e.g., `fly status`, `kubectl rollout status`)
- C) Check the GitHub Actions workflow status
- D) No automated way — just check the URL loads
4. **Any pre-merge or post-merge hooks?**
- Commands to run before merging (e.g., `bun run build`)
- Commands to run after merge but before deploy verification
### Step 4: Write configuration
Read CLAUDE.md (or create it). Find and replace the `## Deploy Configuration` section
if it exists, or append it at the end.
```markdown
## Deploy Configuration (configured by /setup-deploy)
- Platform: {platform}
- Production URL: {url}
- Deploy workflow: {workflow file or "auto-deploy on push"}
- Deploy status command: {command or "HTTP health check"}
- Merge method: {squash/merge/rebase}
- Project type: {web app / API / CLI / library}
- Post-deploy health check: {health check URL or command}
### Custom deploy hooks
- Pre-merge: {command or "none"}
- Deploy trigger: {command or "automatic on push to main"}
- Deploy status: {command or "poll production URL"}
- Health check: {URL or command}
```
### Step 5: Verify
After writing, verify the configuration works:
1. If a health check URL was configured, try it:
```bash
curl -sf "{health-check-url}" -o /dev/null -w "%{http_code}" 2>/dev/null || echo "UNREACHABLE"
```
2. If a deploy status command was configured, try it:
```bash
{deploy-status-command} 2>/dev/null | head -5 || echo "COMMAND_FAILED"
```
Report results. If anything failed, note it but don't block — the config is still
useful even if the health check is temporarily unreachable.
### Step 6: Summary
```
DEPLOY CONFIGURATION — COMPLETE
════════════════════════════════
Platform: {platform}
URL: {url}
Health check: {health check}
Status cmd: {status command}
Merge method: {merge method}
Saved to CLAUDE.md. /land-and-deploy will use these settings automatically.
Next steps:
- Run /land-and-deploy to merge and deploy your current PR
- Edit the "## Deploy Configuration" section in CLAUDE.md to change settings
- Run /setup-deploy again to reconfigure
```
## Important Rules
- **Never expose secrets.** Don't print full API keys, tokens, or passwords.
- **Confirm with the user.** Always show the detected config and ask for confirmation before writing.
- **CLAUDE.md is the source of truth.** All configuration lives there — not in a separate config file.
- **Idempotent.** Running /setup-deploy multiple times overwrites the previous config cleanly.
- **Platform CLIs are optional.** If `fly` or `vercel` CLI isn't installed, fall back to URL-based health checks.

220
skills/setup-deploy/SKILL.md.tmpl
View File

@ -1,220 +0,0 @@
---
name: setup-deploy
version: 1.0.0
description: |
Configure deployment settings for /land-and-deploy. Detects your deploy
platform (Fly.io, Render, Vercel, Netlify, Heroku, GitHub Actions, custom),
production URL, health check endpoints, and deploy status commands. Writes
the configuration to CLAUDE.md so all future deploys are automatic.
Use when: "setup deploy", "configure deployment", "set up land-and-deploy",
"how do I deploy with gstack", "add deploy config".
allowed-tools:
- Bash
- Read
- Write
- Edit
- Glob
- Grep
- AskUserQuestion
---
{{PREAMBLE}}
# /setup-deploy — Configure Deployment for gstack
You are helping the user configure their deployment so `/land-and-deploy` works
automatically. Your job is to detect the deploy platform, production URL, health
checks, and deploy status commands — then persist everything to CLAUDE.md.
After this runs once, `/land-and-deploy` reads CLAUDE.md and skips detection entirely.
## User-invocable
When the user types `/setup-deploy`, run this skill.
## Instructions
### Step 1: Check existing configuration
```bash
grep -A 20 "## Deploy Configuration" CLAUDE.md 2>/dev/null || echo "NO_CONFIG"
```
If configuration already exists, show it and ask:
- **Context:** Deploy configuration already exists in CLAUDE.md.
- **RECOMMENDATION:** Choose A to update if your setup changed.
- A) Reconfigure from scratch (overwrite existing)
- B) Edit specific fields (show current config, let me change one thing)
- C) Done — configuration looks correct
If the user picks C, stop.
### Step 2: Detect platform
Run the platform detection from the deploy bootstrap:
```bash
# Platform config files
[ -f fly.toml ] && echo "PLATFORM:fly" && cat fly.toml
[ -f render.yaml ] && echo "PLATFORM:render" && cat render.yaml
[ -f vercel.json ] || [ -d .vercel ] && echo "PLATFORM:vercel"
[ -f netlify.toml ] && echo "PLATFORM:netlify" && cat netlify.toml
[ -f Procfile ] && echo "PLATFORM:heroku"
[ -f railway.json ] || [ -f railway.toml ] && echo "PLATFORM:railway"
# GitHub Actions deploy workflows
for f in .github/workflows/*.yml .github/workflows/*.yaml; do
[ -f "$f" ] && grep -qiE "deploy|release|production|staging|cd" "$f" 2>/dev/null && echo "DEPLOY_WORKFLOW:$f"
done
# Project type
[ -f package.json ] && grep -q '"bin"' package.json 2>/dev/null && echo "PROJECT_TYPE:cli"
ls *.gemspec 2>/dev/null && echo "PROJECT_TYPE:library"
```
### Step 3: Platform-specific setup
Based on what was detected, guide the user through platform-specific configuration.
#### Fly.io
If `fly.toml` detected:
1. Extract app name: `grep -m1 "^app" fly.toml | sed 's/app = "\(.*\)"/\1/'`
2. Check if `fly` CLI is installed: `which fly 2>/dev/null`
3. If installed, verify: `fly status --app {app} 2>/dev/null`
4. Infer URL: `https://{app}.fly.dev`
5. Set deploy status command: `fly status --app {app}`
6. Set health check: `https://{app}.fly.dev` (or `/health` if the app has one)
Ask the user to confirm the production URL. Some Fly apps use custom domains.
#### Render
If `render.yaml` detected:
1. Extract service name and type from render.yaml
2. Check for Render API key: `echo $RENDER_API_KEY | head -c 4` (don't expose the full key)
3. Infer URL: `https://{service-name}.onrender.com`
4. Render deploys automatically on push to the connected branch — no deploy workflow needed
5. Set health check: the inferred URL
Ask the user to confirm. Render uses auto-deploy from the connected git branch — after
merge to main, Render picks it up automatically. The "deploy wait" in /land-and-deploy
should poll the Render URL until it responds with the new version.
#### Vercel
If vercel.json or .vercel detected:
1. Check for `vercel` CLI: `which vercel 2>/dev/null`
2. If installed: `vercel ls --prod 2>/dev/null | head -3`
3. Vercel deploys automatically on push — preview on PR, production on merge to main
4. Set health check: the production URL from vercel project settings
#### Netlify
If netlify.toml detected:
1. Extract site info from netlify.toml
2. Netlify deploys automatically on push
3. Set health check: the production URL
#### GitHub Actions only
If deploy workflows detected but no platform config:
1. Read the workflow file to understand what it does
2. Extract the deploy target (if mentioned)
3. Ask the user for the production URL
#### Custom / Manual
If nothing detected:
Use AskUserQuestion to gather the information:
1. **How are deploys triggered?**
- A) Automatically on push to main (Fly, Render, Vercel, Netlify, etc.)
- B) Via GitHub Actions workflow
- C) Via a deploy script or CLI command (describe it)
- D) Manually (SSH, dashboard, etc.)
- E) This project doesn't deploy (library, CLI, tool)
2. **What's the production URL?** (Free text — the URL where the app runs)
3. **How can gstack check if a deploy succeeded?**
- A) HTTP health check at a specific URL (e.g., /health, /api/status)
- B) CLI command (e.g., `fly status`, `kubectl rollout status`)
- C) Check the GitHub Actions workflow status
- D) No automated way — just check the URL loads
4. **Any pre-merge or post-merge hooks?**
- Commands to run before merging (e.g., `bun run build`)
- Commands to run after merge but before deploy verification
### Step 4: Write configuration
Read CLAUDE.md (or create it). Find and replace the `## Deploy Configuration` section
if it exists, or append it at the end.
```markdown
## Deploy Configuration (configured by /setup-deploy)
- Platform: {platform}
- Production URL: {url}
- Deploy workflow: {workflow file or "auto-deploy on push"}
- Deploy status command: {command or "HTTP health check"}
- Merge method: {squash/merge/rebase}
- Project type: {web app / API / CLI / library}
- Post-deploy health check: {health check URL or command}
### Custom deploy hooks
- Pre-merge: {command or "none"}
- Deploy trigger: {command or "automatic on push to main"}
- Deploy status: {command or "poll production URL"}
- Health check: {URL or command}
```
### Step 5: Verify
After writing, verify the configuration works:
1. If a health check URL was configured, try it:
```bash
curl -sf "{health-check-url}" -o /dev/null -w "%{http_code}" 2>/dev/null || echo "UNREACHABLE"
```
2. If a deploy status command was configured, try it:
```bash
{deploy-status-command} 2>/dev/null | head -5 || echo "COMMAND_FAILED"
```
Report results. If anything failed, note it but don't block — the config is still
useful even if the health check is temporarily unreachable.
### Step 6: Summary
```
DEPLOY CONFIGURATION — COMPLETE
════════════════════════════════
Platform: {platform}
URL: {url}
Health check: {health check}
Status cmd: {status command}
Merge method: {merge method}
Saved to CLAUDE.md. /land-and-deploy will use these settings automatically.
Next steps:
- Run /land-and-deploy to merge and deploy your current PR
- Edit the "## Deploy Configuration" section in CLAUDE.md to change settings
- Run /setup-deploy again to reconfigure
```
## Important Rules
- **Never expose secrets.** Don't print full API keys, tokens, or passwords.
- **Confirm with the user.** Always show the detected config and ask for confirmation before writing.
- **CLAUDE.md is the source of truth.** All configuration lives there — not in a separate config file.
- **Idempotent.** Running /setup-deploy multiple times overwrites the previous config cleanly.
- **Platform CLIs are optional.** If `fly` or `vercel` CLI isn't installed, fall back to URL-based health checks.

82
skills/swift-expert/SKILL.md
View File

@ -1,82 +0,0 @@
---
name: swift-expert
description: >
Swift/iOS 深度专家。当用户需要 Swift 5.9+/SwiftUI 开发、async/await 并发、Core Data、网络层架构、Swift 测试、App 性能优化，或说 "Swift"、"SwiftUI"、"iOS原生" 时使用此技能。
allowed-tools: Read, Glob, Grep, Edit, Write, Bash
maturity: imported
last-reviewed: 2026-03-03
composable: true
enhances: [mobile-expert]
---
# Swift Expert
Senior Swift developer with mastery of Swift 5.9+, Apple's development ecosystem, SwiftUI, async/await concurrency, and protocol-oriented programming.
## Role Definition
You are a senior Swift engineer with 10+ years of Apple platform development. You specialize in Swift 5.9+, SwiftUI, async/await concurrency, protocol-oriented design, and server-side Swift. You build type-safe, performant applications following Apple's API design guidelines.
## When to Use This Skill
- Building iOS/macOS/watchOS/tvOS applications
- Implementing SwiftUI interfaces and state management
- Setting up async/await concurrency and actors
- Creating protocol-oriented architectures
- Optimizing memory and performance
- Integrating UIKit with SwiftUI
## Core Workflow
1. **Architecture Analysis** - Identify platform targets, dependencies, design patterns
2. **Design Protocols** - Create protocol-first APIs with associated types
3. **Implement** - Write type-safe code with async/await and value semantics
4. **Optimize** - Profile with Instruments, ensure thread safety
5. **Test** - Write comprehensive tests with XCTest and async patterns
## Reference Guide
Load detailed guidance based on context:
| Topic | Reference | Load When |
|-------|-----------|-----------|
| SwiftUI | `references/swiftui-patterns.md` | Building views, state management, modifiers |
| Concurrency | `references/async-concurrency.md` | async/await, actors, structured concurrency |
| Protocols | `references/protocol-oriented.md` | Protocol design, generics, type erasure |
| Memory | `references/memory-performance.md` | ARC, weak/unowned, performance optimization |
| Testing | `references/testing-patterns.md` | XCTest, async tests, mocking strategies |
## Constraints
### MUST DO
- Use type hints and inference appropriately
- Follow Swift API Design Guidelines
- Use async/await for asynchronous operations
- Ensure Sendable compliance for concurrency
- Use value types (struct/enum) by default
- Document APIs with markup comments
- Use property wrappers for cross-cutting concerns
- Profile with Instruments before optimizing
### MUST NOT DO
- Use force unwrapping (!) without justification
- Create retain cycles in closures
- Mix synchronous and asynchronous code improperly
- Ignore actor isolation warnings
- Use implicitly unwrapped optionals unnecessarily
- Skip error handling
- Use Objective-C patterns when Swift alternatives exist
- Hardcode platform-specific values
## Output Templates
When implementing Swift features, provide:
1. Protocol definitions and type aliases
2. Model types (structs/classes with value semantics)
3. View implementations (SwiftUI) or view controllers
4. Tests demonstrating usage
5. Brief explanation of architectural decisions
## Knowledge Reference
Swift 5.9+, SwiftUI, UIKit, async/await, actors, structured concurrency, Combine, property wrappers, result builders, protocol-oriented programming, generics, type erasure, ARC, Instruments, XCTest, Swift Package Manager, Vapor

360
skills/swift-expert/references/async-concurrency.md
View File

@ -1,360 +0,0 @@
# Async/Await Concurrency
## Async/Await Basics
```swift
// Async function
func fetchUser(id: Int) async throws -> User {
let url = URL(string: "https://api.example.com/users/\(id)")!
let (data, _) = try await URLSession.shared.data(from: url)
return try JSONDecoder().decode(User.self, from: data)
}
// Calling async functions
func loadUserData() async {
do {
let user = try await fetchUser(id: 123)
print("Loaded: \(user.name)")
} catch {
print("Error: \(error)")
}
}
// Multiple concurrent operations
func fetchMultipleUsers(ids: [Int]) async throws -> [User] {
try await withThrowingTaskGroup(of: User.self) { group in
for id in ids {
group.addTask {
try await fetchUser(id: id)
}
}
var users: [User] = []
for try await user in group {
users.append(user)
}
return users
}
}
```
## Actors
```swift
// Actor for thread-safe state management
actor UserCache {
private var cache: [Int: User] = [:]
private var inProgress: [Int: Task<User, Error>] = [:]
func user(id: Int) async throws -> User {
// Check cache first
if let cached = cache[id] {
return cached
}
// Check if already loading
if let task = inProgress[id] {
return try await task.value
}
// Start new load
let task = Task {
try await fetchUser(id: id)
}
inProgress[id] = task
do {
let user = try await task.value
cache[id] = user
inProgress.removeValue(forKey: id)
return user
} catch {
inProgress.removeValue(forKey: id)
throw error
}
}
func clearCache() {
cache.removeAll()
}
}
// Usage
let cache = UserCache()
let user = try await cache.user(id: 123)
```
## MainActor
```swift
// UI updates must happen on main thread
@MainActor
class ViewModel: ObservableObject {
@Published var users: [User] = []
@Published var isLoading = false
func loadUsers() async {
isLoading = true
defer { isLoading = false }
do {
// This async work happens off main thread
let loadedUsers = try await fetchMultipleUsers(ids: [1, 2, 3])
// Property updates happen on main thread automatically
users = loadedUsers
} catch {
print("Error: \(error)")
}
}
}
// Isolated functions
@MainActor
func updateUI() {
// This always runs on main thread
}
// Non-isolated functions in MainActor type
@MainActor
class DataManager {
var data: [String] = []
// Runs on main thread
func updateData(_ newData: [String]) {
data = newData
}
// Can run on any thread
nonisolated func processData(_ input: String) -> String {
return input.uppercased()
}
}
```
## Structured Concurrency
```swift
// Task groups for dynamic concurrency
func downloadImages(urls: [URL]) async throws -> [UIImage] {
try await withThrowingTaskGroup(of: (Int, UIImage).self) { group in
for (index, url) in urls.enumerated() {
group.addTask {
let (data, _) = try await URLSession.shared.data(from: url)
guard let image = UIImage(data: data) else {
throw ImageError.invalidData
}
return (index, image)
}
}
var images = [UIImage?](repeating: nil, count: urls.count)
for try await (index, image) in group {
images[index] = image
}
return images.compactMap { $0 }
}
}
// Parallel async-let
func loadDashboard() async throws -> Dashboard {
async let user = fetchUser(id: currentUserID)
async let posts = fetchPosts()
async let notifications = fetchNotifications()
return try await Dashboard(
user: user,
posts: posts,
notifications: notifications
)
}
```
## Task Management
```swift
// Detached tasks
func backgroundWork() {
Task.detached(priority: .background) {
// Runs independently, doesn't inherit context
await performHeavyComputation()
}
}
// Cancellation
class DataLoader {
private var loadTask: Task<Void, Never>?
func startLoading() {
loadTask?.cancel()
loadTask = Task {
do {
for try await item in itemStream() {
// Check for cancellation
try Task.checkCancellation()
await process(item)
// Alternative cancellation check
if Task.isCancelled {
break
}
}
} catch is CancellationError {
print("Task cancelled")
} catch {
print("Error: \(error)")
}
}
}
func stopLoading() {
loadTask?.cancel()
loadTask = nil
}
}
// Task priorities
Task(priority: .high) {
await criticalWork()
}
Task(priority: .low) {
await backgroundWork()
}
```
## AsyncSequence
```swift
// Custom AsyncSequence
struct NumberSequence: AsyncSequence {
typealias Element = Int
let range: Range<Int>
struct AsyncIterator: AsyncIteratorProtocol {
var current: Int
let end: Int
mutating func next() async -> Int? {
guard current < end else { return nil }
// Simulate async work
try? await Task.sleep(for: .milliseconds(100))
defer { current += 1 }
return current
}
}
func makeAsyncIterator() -> AsyncIterator {
AsyncIterator(current: range.lowerBound, end: range.upperBound)
}
}
// Usage
for await number in NumberSequence(range: 0..<10) {
print(number)
}
// Async stream
func eventStream() -> AsyncStream<Event> {
AsyncStream { continuation in
let observer = NotificationCenter.default.addObserver(
forName: .eventOccurred,
object: nil,
queue: nil
) { notification in
if let event = notification.object as? Event {
continuation.yield(event)
}
}
continuation.onTermination = { _ in
NotificationCenter.default.removeObserver(observer)
}
}
}
```
## Sendable Protocol
```swift
// Sendable types can be safely passed across concurrency domains
struct User: Sendable {
let id: Int
let name: String
}
// Non-Sendable by default (has mutable state)
class ViewModel {
var data: [String] = []
}
// Make it Sendable with @unchecked (use carefully!)
class SafeViewModel: @unchecked Sendable {
private let lock = NSLock()
private var _data: [String] = []
var data: [String] {
lock.lock()
defer { lock.unlock() }
return _data
}
func setData(_ newData: [String]) {
lock.lock()
defer { lock.unlock() }
_data = newData
}
}
// Generic with Sendable constraint
func processData<T: Sendable>(_ data: T) async -> T {
// Can safely pass data across concurrency boundaries
await Task.detached {
return data
}.value
}
```
## Continuations
```swift
// Bridging callback-based APIs to async/await
func fetchDataAsync() async throws -> Data {
try await withCheckedThrowingContinuation { continuation in
fetchDataWithCallback { result in
switch result {
case .success(let data):
continuation.resume(returning: data)
case .failure(let error):
continuation.resume(throwing: error)
}
}
}
}
// Unsafe continuations for performance-critical code
func unsafeFetchDataAsync() async -> Data {
await withUnsafeContinuation { continuation in
fetchDataWithCallback { data in
continuation.resume(returning: data)
}
}
}
```
## Best Practices
- Use actors for mutable shared state
- Prefer async/await over completion handlers
- Use MainActor for UI-related code
- Leverage structured concurrency (task groups, async-let)
- Check for cancellation in long-running tasks
- Mark types as Sendable when safe
- Use continuations to bridge legacy async code
- Avoid blocking in async contexts
- Use Task.detached sparingly (breaks structured concurrency)

377
skills/swift-expert/references/memory-performance.md
View File

@ -1,377 +0,0 @@
# Memory & Performance
## Automatic Reference Counting (ARC)
```swift
// Strong references (default)
class Person {
let name: String
var apartment: Apartment?
init(name: String) {
self.name = name
}
deinit {
print("\(name) is being deinitialized")
}
}
class Apartment {
let unit: String
weak var tenant: Person? // Weak to break retain cycle
init(unit: String) {
self.unit = unit
}
deinit {
print("Apartment \(unit) is being deinitialized")
}
}
var john: Person? = Person(name: "John")
var unit4A: Apartment? = Apartment(unit: "4A")
john?.apartment = unit4A
unit4A?.tenant = john
// Setting to nil will properly deallocate both
john = nil
unit4A = nil
```
## Weak and Unowned References
```swift
// Weak - optional reference that doesn't keep object alive
class ViewController: UIViewController {
weak var delegate: ViewControllerDelegate?
func performAction() {
delegate?.didPerformAction()
}
}
// Unowned - non-optional reference, assumes target outlives owner
class Customer {
let name: String
var card: CreditCard?
init(name: String) {
self.name = name
}
}
class CreditCard {
let number: String
unowned let customer: Customer // Customer always outlives card
init(number: String, customer: Customer) {
self.number = number
self.customer = customer
}
}
// Unowned optional (Swift 5+)
class Department {
var courses: [Course] = []
}
class Course {
unowned var department: Department
unowned var nextCourse: Course?
init(department: Department) {
self.department = department
}
}
```
## Capture Lists in Closures
```swift
class DataManager {
var data: [String] = []
func loadData() {
// Strong reference cycle - DataManager won't be deallocated
NetworkManager.fetch { response in
self.data = response // self is captured strongly
}
// Weak self - breaks cycle
NetworkManager.fetch { [weak self] response in
guard let self = self else { return }
self.data = response
}
// Unowned self - when self definitely outlives closure
NetworkManager.fetch { [unowned self] response in
self.data = response // Crashes if self is deallocated
}
// Capturing specific values
let identifier = UUID()
NetworkManager.fetch { [identifier] response in
print("Request \(identifier) completed")
}
}
}
```
## Value Semantics
```swift
// Structs provide automatic copy-on-write for collections
struct User {
var name: String
var friends: [String] // Copy-on-write
}
var user1 = User(name: "Alice", friends: ["Bob"])
var user2 = user1 // Shallow copy
user2.friends.append("Charlie") // Now triggers deep copy
print(user1.friends) // ["Bob"]
print(user2.friends) // ["Bob", "Charlie"]
// Custom copy-on-write
final class Storage<T> {
var value: T
init(_ value: T) { self.value = value }
}
struct MyArray<Element> {
private var storage: Storage<[Element]>
init(_ elements: [Element] = []) {
storage = Storage(elements)
}
var value: [Element] {
get { storage.value }
set {
if !isKnownUniquelyReferenced(&storage) {
storage = Storage(newValue)
} else {
storage.value = newValue
}
}
}
mutating func append(_ element: Element) {
if !isKnownUniquelyReferenced(&storage) {
storage = Storage(storage.value)
}
storage.value.append(element)
}
}
```
## Performance Optimization
```swift
// Use lazy properties for expensive computations
class Report {
let data: [DataPoint]
lazy var summary: String = {
// Expensive computation only when accessed
data.map { $0.description }.joined(separator: "\n")
}()
init(data: [DataPoint]) {
self.data = data
}
}
// Avoid repeated type casting
// Bad
for item in items {
if let user = item as? User {
processUser(user)
}
}
// Good
let users = items.compactMap { $0 as? User }
for user in users {
processUser(user)
}
// Use contiguous storage
// Slower - pointer indirection for each element
let arrayOfClasses: [MyClass] = [MyClass(), MyClass()]
// Faster - contiguous memory
let arrayOfStructs: [MyStruct] = [MyStruct(), MyStruct()]
// Avoid string concatenation in loops
// Bad
var result = ""
for item in items {
result += item.description // Allocates new string each time
}
// Good
let result = items.map { $0.description }.joined()
// Or
var result = ""
result.reserveCapacity(estimatedSize)
for item in items {
result.append(item.description)
}
```
## Collection Performance
```swift
// Choose the right collection type
// Array - ordered, random access O(1), append O(1) amortized
let ordered: [Int] = [1, 2, 3]
// Set - unique elements, contains O(1), no order
let unique: Set<Int> = [1, 2, 3]
// Dictionary - key-value pairs, lookup O(1)
let mapping: [String: Int] = ["a": 1, "b": 2]
// Use ContiguousArray for performance-critical code
let contiguous = ContiguousArray<MyStruct>(repeating: MyStruct(), count: 1000)
// Reserve capacity for known sizes
var numbers: [Int] = []
numbers.reserveCapacity(1000)
for i in 0..<1000 {
numbers.append(i)
}
// Use enumerated() instead of indices
// Bad
for i in 0..<array.count {
process(index: i, value: array[i])
}
// Good
for (index, value) in array.enumerated() {
process(index: index, value: value)
}
```
## Memory Profiling with Instruments
```swift
// Add markers for profiling
import os.signpost
let log = OSLog(subsystem: "com.example.app", category: "Performance")
func processData() {
os_signpost(.begin, log: log, name: "Data Processing")
defer { os_signpost(.end, log: log, name: "Data Processing") }
// Processing code
}
// Autoreleasepool for memory-intensive loops
func processLargeDataset() {
for batch in dataBatches {
autoreleasepool {
// Process batch
// Memory released at end of each iteration
}
}
}
// Check for memory leaks
#if DEBUG
extension NSObject {
static func trackAllocations() {
let count = performSelector(
Selector(("instancesRespond:"))
)
print("\(self): \(count) instances")
}
}
#endif
```
## Optimization Levels
```swift
// Whole Module Optimization in Package.swift
let package = Package(
name: "MyApp",
products: [
.executable(name: "MyApp", targets: ["MyApp"])
],
targets: [
.target(
name: "MyApp",
swiftSettings: [
.unsafeFlags(["-O"], .when(configuration: .release))
]
)
]
)
// Inline optimization
@inline(__always)
func criticalPath() {
// Always inlined
}
@inline(never)
func debugHelper() {
// Never inlined, good for debugging
}
// Optimization attributes
@_specialize(where T == Int)
@_specialize(where T == String)
func process<T>(_ value: T) {
// Specialized versions generated
}
```
## Memory Warnings
```swift
class ImageCache {
private var cache: [String: UIImage] = [:]
init() {
NotificationCenter.default.addObserver(
self,
selector: #selector(clearCache),
name: UIApplication.didReceiveMemoryWarningNotification,
object: nil
)
}
@objc private func clearCache() {
cache.removeAll()
}
deinit {
NotificationCenter.default.removeObserver(self)
}
}
```
## Best Practices
- Use value types (structs) by default
- Use weak references for delegates
- Use unowned when lifetime is guaranteed
- Always use capture lists in closures that reference self
- Profile before optimizing (use Instruments)
- Reserve collection capacity when size is known
- Use lazy properties for expensive computations
- Implement copy-on-write for custom types with reference storage
- Handle memory warnings in iOS apps
- Use autoreleasepool for memory-intensive loops
- Choose appropriate collection types
- Avoid premature optimization - measure first

354
skills/swift-expert/references/protocol-oriented.md
View File

@ -1,354 +0,0 @@
# Protocol-Oriented Programming
## Protocol Basics
```swift
// Protocol with requirements
protocol Drawable {
var boundingBox: CGRect { get }
func draw(in context: CGContext)
}
// Protocol with default implementation
extension Drawable {
func draw(in context: CGContext) {
// Default drawing behavior
context.stroke(boundingBox)
}
}
// Struct conforming to protocol
struct Circle: Drawable {
let center: CGPoint
let radius: CGFloat
var boundingBox: CGRect {
CGRect(
x: center.x - radius,
y: center.y - radius,
width: radius * 2,
height: radius * 2
)
}
}
```
## Associated Types
```swift
// Protocol with associated type
protocol Container {
associatedtype Item
var count: Int { get }
mutating func append(_ item: Item)
subscript(index: Int) -> Item { get }
}
// Generic struct conforming
struct Stack<Element>: Container {
typealias Item = Element // Can be inferred
private var items: [Element] = []
var count: Int { items.count }
mutating func append(_ item: Element) {
items.append(item)
}
subscript(index: Int) -> Element {
items[index]
}
}
// Using where clause with associated types
extension Container where Item: Equatable {
func firstIndex(of item: Item) -> Int? {
for (index, current) in enumerated() where current == item {
return index
}
return nil
}
}
```
## Protocol Composition
```swift
// Multiple protocol conformance
protocol Named {
var name: String { get }
}
protocol Aged {
var age: Int { get }
}
// Composing protocols
typealias Person = Named & Aged
func greet(_ person: some Named & Aged) {
print("Hello \(person.name), age \(person.age)")
}
// Protocol composition in constraints
func process<T: Codable & Hashable>(_ items: [T]) {
// T must conform to both Codable and Hashable
}
```
## Generics with Protocols
```swift
// Generic function with protocol constraint
func compare<T: Comparable>(_ a: T, _ b: T) -> T {
return a > b ? a : b
}
// Generic type with protocol constraint
class Repository<Model: Codable & Identifiable> {
private var items: [Model.ID: Model] = [:]
func save(_ model: Model) {
items[model.id] = model
}
func find(id: Model.ID) -> Model? {
items[id]
}
func all() -> [Model] {
Array(items.values)
}
}
// Using opaque return types
func makeCollection() -> some Collection {
return [1, 2, 3, 4, 5]
}
// Primary associated types (Swift 5.7+)
protocol DataSource<Element> {
associatedtype Element
func fetch() async throws -> [Element]
}
func loadData<T>(from source: some DataSource<T>) async throws -> [T] {
try await source.fetch()
}
```
## Type Erasure
```swift
// Problem: Can't use protocol with associated types as type
// protocol Storage {
// associatedtype Item
// func store(_ item: Item)
// }
// var storage: Storage // Error: protocol can only be used as constraint
// Solution: Type-erased wrapper
protocol Storage {
associatedtype Item
func store(_ item: Item)
func retrieve() -> Item?
}
struct AnyStorage<T>: Storage {
typealias Item = T
private let _store: (T) -> Void
private let _retrieve: () -> T?
init<S: Storage>(_ storage: S) where S.Item == T {
_store = storage.store
_retrieve = storage.retrieve
}
func store(_ item: T) {
_store(item)
}
func retrieve() -> T? {
_retrieve()
}
}
// Now we can use it as a type
class MemoryStorage<T>: Storage {
private var item: T?
func store(_ item: T) {
self.item = item
}
func retrieve() -> T? {
item
}
}
let storage: AnyStorage<String> = AnyStorage(MemoryStorage<String>())
```
## Protocol Inheritance
```swift
// Protocol inheriting from another
protocol Identifiable {
var id: UUID { get }
}
protocol Timestampable {
var createdAt: Date { get }
var updatedAt: Date { get }
}
protocol Entity: Identifiable, Timestampable {
var version: Int { get }
}
struct User: Entity {
let id: UUID
let createdAt: Date
var updatedAt: Date
var version: Int
var name: String
}
```
## Conditional Conformance
```swift
// Make Array conform to protocol when elements conform
protocol Summarizable {
var summary: String { get }
}
extension Array: Summarizable where Element: Summarizable {
var summary: String {
map { $0.summary }.joined(separator: ", ")
}
}
struct Task: Summarizable {
let title: String
var summary: String { title }
}
let tasks = [Task(title: "Buy milk"), Task(title: "Walk dog")]
print(tasks.summary) // "Buy milk, Walk dog"
```
## Protocol Extensions
```swift
// Adding functionality to all conforming types
protocol Collection {
associatedtype Element
var count: Int { get }
subscript(index: Int) -> Element { get }
}
extension Collection {
var isEmpty: Bool {
count == 0
}
func map<T>(_ transform: (Element) -> T) -> [T] {
var result: [T] = []
for i in 0..<count {
result.append(transform(self[i]))
}
return result
}
}
// Constrained extensions
extension Collection where Element: Numeric {
func sum() -> Element {
var total: Element = 0
for i in 0..<count {
total += self[i]
}
return total
}
}
```
## Advanced Patterns
```swift
// Phantom types for type safety
enum Celsius {}
enum Fahrenheit {}
struct Temperature<Unit> {
let value: Double
init(_ value: Double) {
self.value = value
}
}
extension Temperature where Unit == Celsius {
func toFahrenheit() -> Temperature<Fahrenheit> {
Temperature<Fahrenheit>(value * 9/5 + 32)
}
}
extension Temperature where Unit == Fahrenheit {
func toCelsius() -> Temperature<Celsius> {
Temperature<Celsius>((value - 32) * 5/9)
}
}
let celsius = Temperature<Celsius>(100)
let fahrenheit = celsius.toFahrenheit()
// Witness tables pattern
protocol Encoder {
func encode<T: Encodable>(_ value: T) throws -> Data
}
protocol Decoder {
func decode<T: Decodable>(_ type: T.Type, from data: Data) throws -> T
}
struct Codec<E: Encoder, D: Decoder> {
let encoder: E
let decoder: D
func roundtrip<T: Codable>(_ value: T) throws -> T {
let data = try encoder.encode(value)
return try decoder.decode(T.self, from: data)
}
}
```
## Retroactive Modeling
```swift
// Adding protocol conformance to types you don't own
extension Int: Identifiable {
public var id: Int { self }
}
// Now Int can be used where Identifiable is required
let numbers: [Int] = [1, 2, 3]
ForEach(numbers) { number in
Text("\(number)")
}
```
## Best Practices
- Prefer protocols over base classes for abstraction
- Use protocol extensions for default implementations
- Design protocols with single responsibility
- Use associated types for generic protocols
- Apply type erasure when needed for storage
- Leverage conditional conformance
- Use opaque return types (some Protocol) for implementation hiding
- Compose small protocols rather than large ones
- Document protocol requirements and guarantees
- Consider protocol inheritance for layered abstraction

291
skills/swift-expert/references/swiftui-patterns.md
View File

@ -1,291 +0,0 @@
# SwiftUI Patterns
## State Management
```swift
import SwiftUI
// @State for local view state
struct CounterView: View {
@State private var count = 0
var body: some View {
VStack {
Text("Count: \(count)")
Button("Increment") { count += 1 }
}
}
}
// @Binding for two-way data flow
struct ToggleView: View {
@Binding var isOn: Bool
var body: some View {
Toggle("Enable Feature", isOn: $isOn)
}
}
// @StateObject for observable objects (view owns it)
class ViewModel: ObservableObject {
@Published var items: [String] = []
@Published var isLoading = false
}
struct ContentView: View {
@StateObject private var viewModel = ViewModel()
var body: some View {
List(viewModel.items, id: \.self) { item in
Text(item)
}
}
}
// @ObservedObject for passed-in observable objects
struct DetailView: View {
@ObservedObject var viewModel: ViewModel
}
// @EnvironmentObject for dependency injection
struct AppView: View {
@EnvironmentObject var appState: AppState
}
```
## Modern View Composition
```swift
// View builder for custom containers
struct ConditionalView<Content: View>: View {
let condition: Bool
@ViewBuilder let content: () -> Content
var body: some View {
if condition {
content()
} else {
EmptyView()
}
}
}
// Custom ViewModifier
struct CardModifier: ViewModifier {
func body(content: Content) -> some View {
content
.padding()
.background(Color.white)
.cornerRadius(12)
.shadow(radius: 4)
}
}
extension View {
func cardStyle() -> some View {
modifier(CardModifier())
}
}
// Usage
Text("Hello")
.cardStyle()
```
## Environment Values
```swift
// Custom environment key
private struct ThemeKey: EnvironmentKey {
static let defaultValue: Theme = .light
}
extension EnvironmentValues {
var theme: Theme {
get { self[ThemeKey.self] }
set { self[ThemeKey.self] = newValue }
}
}
extension View {
func theme(_ theme: Theme) -> some View {
environment(\.theme, theme)
}
}
// Usage
struct ThemedView: View {
@Environment(\.theme) var theme
var body: some View {
Text("Themed")
.foregroundColor(theme.textColor)
}
}
```
## Preference Keys
```swift
// Collecting data from child views
struct SizePreferenceKey: PreferenceKey {
static var defaultValue: CGSize = .zero
static func reduce(value: inout CGSize, nextValue: () -> CGSize) {
value = nextValue()
}
}
struct MeasurableView: View {
@State private var size: CGSize = .zero
var body: some View {
Text("Measure me")
.background(
GeometryReader { geometry in
Color.clear
.preference(key: SizePreferenceKey.self, value: geometry.size)
}
)
.onPreferenceChange(SizePreferenceKey.self) { newSize in
size = newSize
}
}
}
```
## Animations
```swift
// Implicit animations
struct AnimatedView: View {
@State private var scale: CGFloat = 1.0
var body: some View {
Circle()
.scaleEffect(scale)
.animation(.spring(response: 0.5, dampingFraction: 0.6), value: scale)
.onTapGesture {
scale = scale == 1.0 ? 1.5 : 1.0
}
}
}
// Explicit animations
struct ExplicitAnimationView: View {
@State private var offset: CGFloat = 0
var body: some View {
Text("Slide")
.offset(x: offset)
.onTapGesture {
withAnimation(.easeInOut(duration: 0.3)) {
offset = offset == 0 ? 100 : 0
}
}
}
}
// Custom transitions
extension AnyTransition {
static var slideAndFade: AnyTransition {
AnyTransition.slide.combined(with: .opacity)
}
}
```
## Async/Await Integration
```swift
struct AsyncDataView: View {
@State private var data: [Item] = []
@State private var isLoading = false
var body: some View {
List(data) { item in
Text(item.title)
}
.task {
await loadData()
}
.refreshable {
await loadData()
}
}
private func loadData() async {
isLoading = true
defer { isLoading = false }
do {
data = try await API.fetchItems()
} catch {
print("Error: \(error)")
}
}
}
```
## Custom Layouts (iOS 16+)
```swift
struct WaterfallLayout: Layout {
var columns: Int = 2
var spacing: CGFloat = 8
func sizeThatFits(
proposal: ProposedViewSize,
subviews: Subviews,
cache: inout ()
) -> CGSize {
// Calculate total size needed
let columnWidth = (proposal.width! - spacing * CGFloat(columns - 1)) / CGFloat(columns)
var columnHeights = Array(repeating: CGFloat(0), count: columns)
for subview in subviews {
let column = columnHeights.enumerated().min(by: { $0.element < $1.element })!.offset
let size = subview.sizeThatFits(.init(width: columnWidth, height: nil))
columnHeights[column] += size.height + spacing
}
return CGSize(
width: proposal.width!,
height: columnHeights.max()! - spacing
)
}
func placeSubviews(
in bounds: CGRect,
proposal: ProposedViewSize,
subviews: Subviews,
cache: inout ()
) {
let columnWidth = (bounds.width - spacing * CGFloat(columns - 1)) / CGFloat(columns)
var columnHeights = Array(repeating: CGFloat(0), count: columns)
for subview in subviews {
let column = columnHeights.enumerated().min(by: { $0.element < $1.element })!.offset
let x = bounds.minX + CGFloat(column) * (columnWidth + spacing)
let y = bounds.minY + columnHeights[column]
subview.place(
at: CGPoint(x: x, y: y),
proposal: .init(width: columnWidth, height: nil)
)
columnHeights[column] += subview.dimensions(in: .init(width: columnWidth, height: nil)).height + spacing
}
}
}
```
## Performance Tips
- Use `@State` for simple value types
- Use `@StateObject` for reference types you create
- Use `@ObservedObject` for reference types passed in
- Prefer `@Environment` over prop drilling
- Use `equatable()` modifier for expensive views
- Leverage `id()` modifier to control view identity
- Use `task(id:)` to cancel and restart async work
- Avoid computing expensive values in body - use `@State` or computed properties

399
skills/swift-expert/references/testing-patterns.md
View File

@ -1,399 +0,0 @@
# Testing Patterns
## XCTest Basics
```swift
import XCTest
@testable import MyApp
final class UserTests: XCTestCase {
var sut: UserManager!
override func setUp() {
super.setUp()
sut = UserManager()
}
override func tearDown() {
sut = nil
super.tearDown()
}
func testUserCreation() {
// Given
let name = "John Doe"
let email = "john@example.com"
// When
let user = sut.createUser(name: name, email: email)
// Then
XCTAssertEqual(user.name, name)
XCTAssertEqual(user.email, email)
XCTAssertNotNil(user.id)
}
func testValidation() throws {
// Unwrapping optionals in tests
let user = try XCTUnwrap(sut.findUser(id: 123))
XCTAssertEqual(user.name, "Test User")
}
}
```
## Async Testing
```swift
final class AsyncTests: XCTestCase {
func testAsyncFunction() async throws {
// Test async/await code directly
let result = try await fetchData()
XCTAssertEqual(result.count, 10)
}
func testAsyncSequence() async throws {
var results: [Int] = []
for try await value in numberStream() {
results.append(value)
if results.count >= 5 {
break
}
}
XCTAssertEqual(results.count, 5)
}
func testWithTimeout() async throws {
// Test with timeout
try await withTimeout(seconds: 5) {
try await longRunningOperation()
}
}
func testConcurrentOperations() async throws {
async let result1 = fetchData(id: 1)
async let result2 = fetchData(id: 2)
let (data1, data2) = try await (result1, result2)
XCTAssertNotNil(data1)
XCTAssertNotNil(data2)
}
}
// Helper for timeout
func withTimeout<T>(
seconds: TimeInterval,
operation: @escaping () async throws -> T
) async throws -> T {
try await withThrowingTaskGroup(of: T.self) { group in
group.addTask {
try await operation()
}
group.addTask {
try await Task.sleep(nanoseconds: UInt64(seconds * 1_000_000_000))
throw TimeoutError()
}
let result = try await group.next()!
group.cancelAll()
return result
}
}
```
## Mocking
```swift
// Protocol for dependency injection
protocol DataService {
func fetch(id: Int) async throws -> Data
func save(_ data: Data) async throws
}
// Production implementation
class APIDataService: DataService {
func fetch(id: Int) async throws -> Data {
// Real API call
}
func save(_ data: Data) async throws {
// Real save operation
}
}
// Mock for testing
class MockDataService: DataService {
var fetchCalled = false
var fetchID: Int?
var fetchResult: Data?
var fetchError: Error?
var saveCalled = false
var savedData: Data?
var saveError: Error?
func fetch(id: Int) async throws -> Data {
fetchCalled = true
fetchID = id
if let error = fetchError {
throw error
}
return fetchResult ?? Data()
}
func save(_ data: Data) async throws {
saveCalled = true
savedData = data
if let error = saveError {
throw error
}
}
}
// Using mock in tests
final class DataManagerTests: XCTestCase {
func testDataFetch() async throws {
// Given
let mockService = MockDataService()
mockService.fetchResult = "test data".data(using: .utf8)
let manager = DataManager(service: mockService)
// When
let result = try await manager.loadData(id: 123)
// Then
XCTAssertTrue(mockService.fetchCalled)
XCTAssertEqual(mockService.fetchID, 123)
XCTAssertNotNil(result)
}
}
```
## Test Doubles
```swift
// Spy - records interactions
class SpyDelegate: UserManagerDelegate {
private(set) var didUpdateUserCalled = false
private(set) var updatedUser: User?
private(set) var callCount = 0
func didUpdateUser(_ user: User) {
didUpdateUserCalled = true
updatedUser = user
callCount += 1
}
}
// Stub - provides predetermined responses
class StubNetworkService: NetworkService {
var stubbedResponse: Result<Data, Error> = .success(Data())
func fetch(url: URL) async throws -> Data {
try stubbedResponse.get()
}
}
// Fake - working implementation with shortcuts
class FakeDatabase: Database {
private var storage: [String: Data] = [:]
func save(key: String, value: Data) {
storage[key] = value
}
func load(key: String) -> Data? {
storage[key]
}
func clear() {
storage.removeAll()
}
}
```
## Performance Testing
```swift
final class PerformanceTests: XCTestCase {
func testSortingPerformance() {
let numbers = (0..<10000).shuffled()
measure {
_ = numbers.sorted()
}
}
func testCustomMetrics() {
let metrics: [XCTMetric] = [
XCTClockMetric(),
XCTCPUMetric(),
XCTMemoryMetric(),
XCTStorageMetric()
]
let options = XCTMeasureOptions()
options.iterationCount = 10
measure(metrics: metrics, options: options) {
performExpensiveOperation()
}
}
}
```
## UI Testing
```swift
final class AppUITests: XCTestCase {
var app: XCUIApplication!
override func setUp() {
super.setUp()
continueAfterFailure = false
app = XCUIApplication()
app.launch()
}
func testLoginFlow() {
// Test UI interactions
let emailField = app.textFields["Email"]
emailField.tap()
emailField.typeText("test@example.com")
let passwordField = app.secureTextFields["Password"]
passwordField.tap()
passwordField.typeText("password123")
app.buttons["Login"].tap()
// Verify navigation
XCTAssertTrue(app.navigationBars["Dashboard"].exists)
}
func testButtonEnabled() {
let button = app.buttons["Submit"]
XCTAssertFalse(button.isEnabled)
app.textFields["Username"].tap()
app.textFields["Username"].typeText("testuser")
XCTAssertTrue(button.isEnabled)
}
}
```
## Testing Actors
```swift
final class ActorTests: XCTestCase {
func testActorIsolation() async throws {
actor Counter {
private var value = 0
func increment() -> Int {
value += 1
return value
}
func reset() {
value = 0
}
}
let counter = Counter()
// Test concurrent access
await withTaskGroup(of: Int.self) { group in
for _ in 0..<100 {
group.addTask {
await counter.increment()
}
}
}
let finalValue = await counter.increment()
XCTAssertEqual(finalValue, 101)
}
}
```
## Snapshot Testing
```swift
import SnapshotTesting
final class ViewSnapshotTests: XCTestCase {
func testButtonAppearance() {
let button = UIButton()
button.setTitle("Tap Me", for: .normal)
button.backgroundColor = .blue
button.frame = CGRect(x: 0, y: 0, width: 200, height: 50)
assertSnapshot(matching: button, as: .image)
}
func testViewControllerLayout() {
let vc = MyViewController()
assertSnapshot(matching: vc, as: .image(on: .iPhone13))
}
func testDarkMode() {
let view = MyView()
assertSnapshot(matching: view, as: .image(traits: .init(userInterfaceStyle: .dark)))
}
}
```
## Test Organization
```swift
// MARK: - Test Cases
extension UserManagerTests {
// MARK: Creation Tests
func testUserCreation() { }
func testUserCreationWithInvalidData() { }
// MARK: Validation Tests
func testEmailValidation() { }
func testPasswordValidation() { }
// MARK: Persistence Tests
func testUserSave() { }
func testUserLoad() { }
}
// MARK: - Test Helpers
extension UserManagerTests {
func makeTestUser() -> User {
User(name: "Test", email: "test@example.com")
}
func setupMockData() {
// Common test setup
}
}
```
## Best Practices
- Use `@testable import` to test internal types
- One assertion concept per test (can have multiple XCTAssert calls)
- Use Given-When-Then pattern for clarity
- Name tests descriptively: `test_methodName_condition_expectedResult`
- Use setUp/tearDown for common test setup
- Prefer dependency injection for testability
- Use protocols to enable mocking
- Test edge cases and error conditions
- Use async/await for testing async code
- Measure performance with XCTest metrics
- Use UI testing for critical user flows
- Mock external dependencies
- Keep tests fast and independent
- Use test doubles appropriately (mock, stub, spy, fake)

378
skills/ui-ux-pro-max/SKILL.md
View File

@ -1,378 +0,0 @@
---
name: ui-ux-pro-max
description: "UI/UX design intelligence. 67 styles, 96 palettes, 57 font pairings, 25 charts, 13 stacks (React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui). Actions: plan, build, create, design, implement, review, fix, improve, optimize, enhance, refactor, check UI/UX code. Projects: website, landing page, dashboard, admin panel, e-commerce, SaaS, portfolio, blog, mobile app, .html, .tsx, .vue, .svelte. Elements: button, modal, navbar, sidebar, card, table, form, chart. Styles: glassmorphism, claymorphism, minimalism, brutalism, neumorphism, bento grid, dark mode, responsive, skeuomorphism, flat design. Topics: color palette, accessibility, animation, layout, typography, font pairing, spacing, hover, shadow, gradient. Integrations: shadcn/ui MCP for component search and examples."
maturity: stable
---
# UI/UX Pro Max - Design Intelligence
Comprehensive design guide for web and mobile applications. Contains 67 styles, 96 color palettes, 57 font pairings, 99 UX guidelines, and 25 chart types across 13 technology stacks. Searchable database with priority-based recommendations.
## When to Apply
Reference these guidelines when:
- Designing new UI components or pages
- Choosing color palettes and typography
- Reviewing code for UX issues
- Building landing pages or dashboards
- Implementing accessibility requirements
## Rule Categories by Priority
| Priority | Category | Impact | Domain |
|----------|----------|--------|--------|
| 1 | Accessibility | CRITICAL | `ux` |
| 2 | Touch & Interaction | CRITICAL | `ux` |
| 3 | Performance | HIGH | `ux` |
| 4 | Layout & Responsive | HIGH | `ux` |
| 5 | Typography & Color | MEDIUM | `typography`, `color` |
| 6 | Animation | MEDIUM | `ux` |
| 7 | Style Selection | MEDIUM | `style`, `product` |
| 8 | Charts & Data | LOW | `chart` |
## Quick Reference
### 1. Accessibility (CRITICAL)
- `color-contrast` - Minimum 4.5:1 ratio for normal text
- `focus-states` - Visible focus rings on interactive elements
- `alt-text` - Descriptive alt text for meaningful images
- `aria-labels` - aria-label for icon-only buttons
- `keyboard-nav` - Tab order matches visual order
- `form-labels` - Use label with for attribute
### 2. Touch & Interaction (CRITICAL)
- `touch-target-size` - Minimum 44x44px touch targets
- `hover-vs-tap` - Use click/tap for primary interactions
- `loading-buttons` - Disable button during async operations
- `error-feedback` - Clear error messages near problem
- `cursor-pointer` - Add cursor-pointer to clickable elements
### 3. Performance (HIGH)
- `image-optimization` - Use WebP, srcset, lazy loading
- `reduced-motion` - Check prefers-reduced-motion
- `content-jumping` - Reserve space for async content
### 4. Layout & Responsive (HIGH)
- `viewport-meta` - width=device-width initial-scale=1
- `readable-font-size` - Minimum 16px body text on mobile
- `horizontal-scroll` - Ensure content fits viewport width
- `z-index-management` - Define z-index scale (10, 20, 30, 50)
### 5. Typography & Color (MEDIUM)
- `line-height` - Use 1.5-1.75 for body text
- `line-length` - Limit to 65-75 characters per line
- `font-pairing` - Match heading/body font personalities
### 6. Animation (MEDIUM)
- `duration-timing` - Use 150-300ms for micro-interactions
- `transform-performance` - Use transform/opacity, not width/height
- `loading-states` - Skeleton screens or spinners
### 7. Style Selection (MEDIUM)
- `style-match` - Match style to product type
- `consistency` - Use same style across all pages
- `no-emoji-icons` - Use SVG icons, not emojis
### 8. Charts & Data (LOW)
- `chart-type` - Match chart type to data type
- `color-guidance` - Use accessible color palettes
- `data-table` - Provide table alternative for accessibility
## How to Use
Search specific domains using the CLI tool below.
---
## Prerequisites
Check if Python is installed:
```bash
python3 --version || python --version
```
If Python is not installed, install it based on user's OS:
**macOS:**
```bash
brew install python3
```
**Ubuntu/Debian:**
```bash
sudo apt update && sudo apt install python3
```
**Windows:**
```powershell
winget install Python.Python.3.12
```
---
## How to Use This Skill
When user requests UI/UX work (design, build, create, implement, review, fix, improve), follow this workflow:
### Step 1: Analyze User Requirements
Extract key information from user request:
- **Product type**: SaaS, e-commerce, portfolio, dashboard, landing page, etc.
- **Style keywords**: minimal, playful, professional, elegant, dark mode, etc.
- **Industry**: healthcare, fintech, gaming, education, etc.
- **Stack**: React, Vue, Next.js, or default to `html-tailwind`
### Step 2: Generate Design System (REQUIRED)
**Always start with `--design-system`** to get comprehensive recommendations with reasoning:
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "<product_type> <industry> <keywords>" --design-system [-p "Project Name"]
```
This command:
1. Searches 5 domains in parallel (product, style, color, landing, typography)
2. Applies reasoning rules from `ui-reasoning.csv` to select best matches
3. Returns complete design system: pattern, style, colors, typography, effects
4. Includes anti-patterns to avoid
**Example:**
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "beauty spa wellness service" --design-system -p "Serenity Spa"
```
### Step 2b: Persist Design System (Master + Overrides Pattern)
To save the design system for hierarchical retrieval across sessions, add `--persist`:
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "<query>" --design-system --persist -p "Project Name"
```
This creates:
- `design-system/MASTER.md` — Global Source of Truth with all design rules
- `design-system/pages/` — Folder for page-specific overrides
**With page-specific override:**
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "<query>" --design-system --persist -p "Project Name" --page "dashboard"
```
This also creates:
- `design-system/pages/dashboard.md` — Page-specific deviations from Master
**How hierarchical retrieval works:**
1. When building a specific page (e.g., "Checkout"), first check `design-system/pages/checkout.md`
2. If the page file exists, its rules **override** the Master file
3. If not, use `design-system/MASTER.md` exclusively
### Step 3: Supplement with Detailed Searches (as needed)
After getting the design system, use domain searches to get additional details:
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "<keyword>" --domain <domain> [-n <max_results>]
```
**When to use detailed searches:**
| Need | Domain | Example |
|------|--------|---------|
| More style options | `style` | `--domain style "glassmorphism dark"` |
| Chart recommendations | `chart` | `--domain chart "real-time dashboard"` |
| UX best practices | `ux` | `--domain ux "animation accessibility"` |
| Alternative fonts | `typography` | `--domain typography "elegant luxury"` |
| Landing structure | `landing` | `--domain landing "hero social-proof"` |
### Step 4: Stack Guidelines (Default: html-tailwind)
Get implementation-specific best practices. If user doesn't specify a stack, **default to `html-tailwind`**.
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "<keyword>" --stack html-tailwind
```
Available stacks: `html-tailwind`, `react`, `nextjs`, `vue`, `svelte`, `swiftui`, `react-native`, `flutter`, `shadcn`, `jetpack-compose`
---
## Search Reference
### Available Domains
| Domain | Use For | Example Keywords |
|--------|---------|------------------|
| `product` | Product type recommendations | SaaS, e-commerce, portfolio, healthcare, beauty, service |
| `style` | UI styles, colors, effects | glassmorphism, minimalism, dark mode, brutalism |
| `typography` | Font pairings, Google Fonts | elegant, playful, professional, modern |
| `color` | Color palettes by product type | saas, ecommerce, healthcare, beauty, fintech, service |
| `landing` | Page structure, CTA strategies | hero, hero-centric, testimonial, pricing, social-proof |
| `chart` | Chart types, library recommendations | trend, comparison, timeline, funnel, pie |
| `ux` | Best practices, anti-patterns | animation, accessibility, z-index, loading |
| `react` | React/Next.js performance | waterfall, bundle, suspense, memo, rerender, cache |
| `web` | Web interface guidelines | aria, focus, keyboard, semantic, virtualize |
| `prompt` | AI prompts, CSS keywords | (style name) |
### Available Stacks
| Stack | Focus |
|-------|-------|
| `html-tailwind` | Tailwind utilities, responsive, a11y (DEFAULT) |
| `react` | State, hooks, performance, patterns |
| `nextjs` | SSR, routing, images, API routes |
| `vue` | Composition API, Pinia, Vue Router |
| `svelte` | Runes, stores, SvelteKit |
| `swiftui` | Views, State, Navigation, Animation |
| `react-native` | Components, Navigation, Lists |
| `flutter` | Widgets, State, Layout, Theming |
| `shadcn` | shadcn/ui components, theming, forms, patterns |
| `jetpack-compose` | Composables, Modifiers, State Hoisting, Recomposition |
---
## Example Workflow
**User request:** "Làm landing page cho dịch vụ chăm sóc da chuyên nghiệp"
### Step 1: Analyze Requirements
- Product type: Beauty/Spa service
- Style keywords: elegant, professional, soft
- Industry: Beauty/Wellness
- Stack: html-tailwind (default)
### Step 2: Generate Design System (REQUIRED)
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "beauty spa wellness service elegant" --design-system -p "Serenity Spa"
```
**Output:** Complete design system with pattern, style, colors, typography, effects, and anti-patterns.
### Step 3: Supplement with Detailed Searches (as needed)
```bash
# Get UX guidelines for animation and accessibility
python3 skills/ui-ux-pro-max/scripts/search.py "animation accessibility" --domain ux
# Get alternative typography options if needed
python3 skills/ui-ux-pro-max/scripts/search.py "elegant luxury serif" --domain typography
```
### Step 4: Stack Guidelines
```bash
python3 skills/ui-ux-pro-max/scripts/search.py "layout responsive form" --stack html-tailwind
```
**Then:** Synthesize design system + detailed searches and implement the design.
---
## Output Formats
The `--design-system` flag supports two output formats:
```bash
# ASCII box (default) - best for terminal display
python3 skills/ui-ux-pro-max/scripts/search.py "fintech crypto" --design-system
# Markdown - best for documentation
python3 skills/ui-ux-pro-max/scripts/search.py "fintech crypto" --design-system -f markdown
```
---
## Tips for Better Results
1. **Be specific with keywords** - "healthcare SaaS dashboard" > "app"
2. **Search multiple times** - Different keywords reveal different insights
3. **Combine domains** - Style + Typography + Color = Complete design system
4. **Always check UX** - Search "animation", "z-index", "accessibility" for common issues
5. **Use stack flag** - Get implementation-specific best practices
6. **Iterate** - If first search doesn't match, try different keywords
---
## Common Rules for Professional UI
These are frequently overlooked issues that make UI look unprofessional:
### Icons & Visual Elements
| Rule | Do | Don't |
|------|----|----- |
| **No emoji icons** | Use SVG icons (Heroicons, Lucide, Simple Icons) | Use emojis like 🎨 🚀 ⚙️ as UI icons |
| **Stable hover states** | Use color/opacity transitions on hover | Use scale transforms that shift layout |
| **Correct brand logos** | Research official SVG from Simple Icons | Guess or use incorrect logo paths |
| **Consistent icon sizing** | Use fixed viewBox (24x24) with w-6 h-6 | Mix different icon sizes randomly |
### Interaction & Cursor
| Rule | Do | Don't |
|------|----|----- |
| **Cursor pointer** | Add `cursor-pointer` to all clickable/hoverable cards | Leave default cursor on interactive elements |
| **Hover feedback** | Provide visual feedback (color, shadow, border) | No indication element is interactive |
| **Smooth transitions** | Use `transition-colors duration-200` | Instant state changes or too slow (>500ms) |
### Light/Dark Mode Contrast
| Rule | Do | Don't |
|------|----|----- |
| **Glass card light mode** | Use `bg-white/80` or higher opacity | Use `bg-white/10` (too transparent) |
| **Text contrast light** | Use `#0F172A` (slate-900) for text | Use `#94A3B8` (slate-400) for body text |
| **Muted text light** | Use `#475569` (slate-600) minimum | Use gray-400 or lighter |
| **Border visibility** | Use `border-gray-200` in light mode | Use `border-white/10` (invisible) |
### Layout & Spacing
| Rule | Do | Don't |
|------|----|----- |
| **Floating navbar** | Add `top-4 left-4 right-4` spacing | Stick navbar to `top-0 left-0 right-0` |
| **Content padding** | Account for fixed navbar height | Let content hide behind fixed elements |
| **Consistent max-width** | Use same `max-w-6xl` or `max-w-7xl` | Mix different container widths |
---
## Pre-Delivery Checklist
Before delivering UI code, verify these items:
### Visual Quality
- [ ] No emojis used as icons (use SVG instead)
- [ ] All icons from consistent icon set (Heroicons/Lucide)
- [ ] Brand logos are correct (verified from Simple Icons)
- [ ] Hover states don't cause layout shift
- [ ] Use theme colors directly (bg-primary) not var() wrapper
### Interaction
- [ ] All clickable elements have `cursor-pointer`
- [ ] Hover states provide clear visual feedback
- [ ] Transitions are smooth (150-300ms)
- [ ] Focus states visible for keyboard navigation
### Light/Dark Mode
- [ ] Light mode text has sufficient contrast (4.5:1 minimum)
- [ ] Glass/transparent elements visible in light mode
- [ ] Borders visible in both modes
- [ ] Test both modes before delivery
### Layout
- [ ] Floating elements have proper spacing from edges
- [ ] No content hidden behind fixed navbars
- [ ] Responsive at 375px, 768px, 1024px, 1440px
- [ ] No horizontal scroll on mobile
### Accessibility
- [ ] All images have alt text
- [ ] Form inputs have labels
- [ ] Color is not the only indicator
- [ ] `prefers-reduced-motion` respected

26
skills/ui-ux-pro-max/data/charts.csv
View File

@ -1,26 +0,0 @@
No,Data Type,Keywords,Best Chart Type,Secondary Options,Color Guidance,Performance Impact,Accessibility Notes,Library Recommendation,Interactive Level
1,Trend Over Time,"trend, time-series, line, growth, timeline, progress",Line Chart,"Area Chart, Smooth Area",Primary: #0080FF. Multiple series: use distinct colors. Fill: 20% opacity,⚡ Excellent (optimized),✓ Clear line patterns for colorblind users. Add pattern overlays.,"Chart.js, Recharts, ApexCharts",Hover + Zoom
2,Compare Categories,"compare, categories, bar, comparison, ranking",Bar Chart (Horizontal or Vertical),"Column Chart, Grouped Bar",Each bar: distinct color. Category: grouped same color. Sorted: descending order,⚡ Excellent,✓ Easy to compare. Add value labels on bars for clarity.,"Chart.js, Recharts, D3.js",Hover + Sort
3,Part-to-Whole,"part-to-whole, pie, donut, percentage, proportion, share",Pie Chart or Donut,"Stacked Bar, Treemap",Colors: 5-6 max. Contrasting palette. Large slices first. Use labels.,⚡ Good (limit 6 slices),⚠ Hard for accessibility. Better: Stacked bar with legend. Avoid pie if >5 items.,"Chart.js, Recharts, D3.js",Hover + Drill
4,Correlation/Distribution,"correlation, distribution, scatter, relationship, pattern",Scatter Plot or Bubble Chart,"Heat Map, Matrix",Color axis: gradient (blue-red). Size: relative. Opacity: 0.6-0.8 to show density,⚠ Moderate (many points),⚠ Provide data table alternative. Use pattern + color distinction.,"D3.js, Plotly, Recharts",Hover + Brush
5,Heatmap/Intensity,"heatmap, heat-map, intensity, density, matrix",Heat Map or Choropleth,"Grid Heat Map, Bubble Heat",Gradient: Cool (blue) to Hot (red). Scale: clear legend. Divergent for ±data,⚡ Excellent (color CSS),⚠ Colorblind: Use pattern overlay. Provide numerical legend.,"D3.js, Plotly, ApexCharts",Hover + Zoom
6,Geographic Data,"geographic, map, location, region, geo, spatial","Choropleth Map, Bubble Map",Geographic Heat Map,Regional: single color gradient or categorized colors. Legend: clear scale,⚠ Moderate (rendering),⚠ Include text labels for regions. Provide data table alternative.,"D3.js, Mapbox, Leaflet",Pan + Zoom + Drill
7,Funnel/Flow,funnel/flow,"Funnel Chart, Sankey",Waterfall (for flows),Stages: gradient (starting color → ending color). Show conversion %,⚡ Good,✓ Clear stage labels + percentages. Good for accessibility if labeled.,"D3.js, Recharts, Custom SVG",Hover + Drill
8,Performance vs Target,performance-vs-target,Gauge Chart or Bullet Chart,"Dial, Thermometer",Performance: Red→Yellow→Green gradient. Target: marker line. Threshold colors,⚡ Good,✓ Add numerical value + percentage label beside gauge.,"D3.js, ApexCharts, Custom SVG",Hover
9,Time-Series Forecast,time-series-forecast,Line with Confidence Band,Ribbon Chart,Actual: solid line #0080FF. Forecast: dashed #FF9500. Band: light shading,⚡ Good,✓ Clearly distinguish actual vs forecast. Add legend.,"Chart.js, ApexCharts, Plotly",Hover + Toggle
10,Anomaly Detection,anomaly-detection,Line Chart with Highlights,Scatter with Alert,Normal: blue #0080FF. Anomaly: red #FF0000 circle/square marker + alert,⚡ Good,✓ Circle/marker for anomalies. Add text alert annotation.,"D3.js, Plotly, ApexCharts",Hover + Alert
11,Hierarchical/Nested Data,hierarchical/nested-data,Treemap,"Sunburst, Nested Donut, Icicle",Parent: distinct hues. Children: lighter shades. White borders 2-3px.,⚠ Moderate,⚠ Poor - provide table alternative. Label large areas.,"D3.js, Recharts, ApexCharts",Hover + Drilldown
12,Flow/Process Data,flow/process-data,Sankey Diagram,"Alluvial, Chord Diagram",Gradient from source to target. Opacity 0.4-0.6 for flows.,⚠ Moderate,⚠ Poor - provide flow table alternative.,"D3.js (d3-sankey), Plotly",Hover + Drilldown
13,Cumulative Changes,cumulative-changes,Waterfall Chart,"Stacked Bar, Cascade",Increases: #4CAF50. Decreases: #F44336. Start: #2196F3. End: #0D47A1.,⚡ Good,✓ Good - clear directional colors with labels.,"ApexCharts, Highcharts, Plotly",Hover
14,Multi-Variable Comparison,multi-variable-comparison,Radar/Spider Chart,"Parallel Coordinates, Grouped Bar",Single: #0080FF 20% fill. Multiple: distinct colors per dataset.,⚡ Good,⚠ Moderate - limit 5-8 axes. Add data table.,"Chart.js, Recharts, ApexCharts",Hover + Toggle
15,Stock/Trading OHLC,stock/trading-ohlc,Candlestick Chart,"OHLC Bar, Heikin-Ashi",Bullish: #26A69A. Bearish: #EF5350. Volume: 40% opacity below.,⚡ Good,⚠ Moderate - provide OHLC data table.,"Lightweight Charts (TradingView), ApexCharts",Real-time + Hover + Zoom
16,Relationship/Connection Data,relationship/connection-data,Network Graph,"Hierarchical Tree, Adjacency Matrix",Node types: categorical colors. Edges: #90A4AE 60% opacity.,❌ Poor (500+ nodes struggles),❌ Very Poor - provide adjacency list alternative.,"D3.js (d3-force), Vis.js, Cytoscape.js",Drilldown + Hover + Drag
17,Distribution/Statistical,distribution/statistical,Box Plot,"Violin Plot, Beeswarm",Box: #BBDEFB. Border: #1976D2. Median: #D32F2F. Outliers: #F44336.,⚡ Excellent,"✓ Good - include stats table (min, Q1, median, Q3, max).","Plotly, D3.js, Chart.js (plugin)",Hover
18,Performance vs Target (Compact),performance-vs-target-(compact),Bullet Chart,"Gauge, Progress Bar","Ranges: #FFCDD2, #FFF9C4, #C8E6C9. Performance: #1976D2. Target: black 3px.",⚡ Excellent,✓ Excellent - compact with clear values.,"D3.js, Plotly, Custom SVG",Hover
19,Proportional/Percentage,proportional/percentage,Waffle Chart,"Pictogram, Stacked Bar 100%",10x10 grid. 3-5 categories max. 2-3px spacing between squares.,⚡ Good,✓ Good - better than pie for accessibility.,"D3.js, React-Waffle, Custom CSS Grid",Hover
20,Hierarchical Proportional,hierarchical-proportional,Sunburst Chart,"Treemap, Icicle, Circle Packing",Center to outer: darker to lighter. 15-20% lighter per level.,⚠ Moderate,⚠ Poor - provide hierarchy table alternative.,"D3.js (d3-hierarchy), Recharts, ApexCharts",Drilldown + Hover
21,Root Cause Analysis,"root cause, decomposition, tree, hierarchy, drill-down, ai-split",Decomposition Tree,"Decision Tree, Flow Chart",Nodes: #2563EB (Primary) vs #EF4444 (Negative impact). Connectors: Neutral grey.,⚠ Moderate (calculation heavy),✓ clear hierarchy. Allow keyboard navigation for nodes.,"Power BI (native), React-Flow, Custom D3.js",Drill + Expand
22,3D Spatial Data,"3d, spatial, immersive, terrain, molecular, volumetric",3D Scatter/Surface Plot,"Volumetric Rendering, Point Cloud",Depth cues: lighting/shading. Z-axis: color gradient (cool to warm).,❌ Heavy (WebGL required),❌ Poor - requires alternative 2D view or data table.,"Three.js, Deck.gl, Plotly 3D",Rotate + Zoom + VR
23,Real-Time Streaming,"streaming, real-time, ticker, live, velocity, pulse",Streaming Area Chart,"Ticker Tape, Moving Gauge",Current: Bright Pulse (#00FF00). History: Fading opacity. Grid: Dark.,⚡ Optimized (canvas/webgl),⚠ Flashing elements - provide pause button. High contrast.,Smoothed D3.js, CanvasJS
24,Sentiment/Emotion,"sentiment, emotion, nlp, opinion, feeling",Word Cloud with Sentiment,"Sentiment Arc, Radar Chart",Positive: #22C55E. Negative: #EF4444. Neutral: #94A3B8. Size = Frequency.,⚡ Good,⚠ Word clouds poor for screen readers. Use list view.,"D3-cloud, Highcharts, Nivo",Hover + Filter
25,Process Mining,"process, mining, variants, path, bottleneck, log",Process Map / Graph,"Directed Acyclic Graph (DAG), Petri Net",Happy path: #10B981 (Thick). Deviations: #F59E0B (Thin). Bottlenecks: #EF4444.,⚠ Moderate to Heavy,⚠ Complex graphs hard to navigate. Provide path summary.,"React-Flow, Cytoscape.js, Recharts",Drag + Node-Click
1 No Data Type Keywords Best Chart Type Secondary Options Color Guidance Performance Impact Accessibility Notes Library Recommendation Interactive Level
2 1 Trend Over Time trend, time-series, line, growth, timeline, progress Line Chart Area Chart, Smooth Area Primary: #0080FF. Multiple series: use distinct colors. Fill: 20% opacity ⚡ Excellent (optimized) ✓ Clear line patterns for colorblind users. Add pattern overlays. Chart.js, Recharts, ApexCharts Hover + Zoom
3 2 Compare Categories compare, categories, bar, comparison, ranking Bar Chart (Horizontal or Vertical) Column Chart, Grouped Bar Each bar: distinct color. Category: grouped same color. Sorted: descending order ⚡ Excellent ✓ Easy to compare. Add value labels on bars for clarity. Chart.js, Recharts, D3.js Hover + Sort
4 3 Part-to-Whole part-to-whole, pie, donut, percentage, proportion, share Pie Chart or Donut Stacked Bar, Treemap Colors: 5-6 max. Contrasting palette. Large slices first. Use labels. ⚡ Good (limit 6 slices) ⚠ Hard for accessibility. Better: Stacked bar with legend. Avoid pie if >5 items. Chart.js, Recharts, D3.js Hover + Drill
5 4 Correlation/Distribution correlation, distribution, scatter, relationship, pattern Scatter Plot or Bubble Chart Heat Map, Matrix Color axis: gradient (blue-red). Size: relative. Opacity: 0.6-0.8 to show density ⚠ Moderate (many points) ⚠ Provide data table alternative. Use pattern + color distinction. D3.js, Plotly, Recharts Hover + Brush
6 5 Heatmap/Intensity heatmap, heat-map, intensity, density, matrix Heat Map or Choropleth Grid Heat Map, Bubble Heat Gradient: Cool (blue) to Hot (red). Scale: clear legend. Divergent for ±data ⚡ Excellent (color CSS) ⚠ Colorblind: Use pattern overlay. Provide numerical legend. D3.js, Plotly, ApexCharts Hover + Zoom
7 6 Geographic Data geographic, map, location, region, geo, spatial Choropleth Map, Bubble Map Geographic Heat Map Regional: single color gradient or categorized colors. Legend: clear scale ⚠ Moderate (rendering) ⚠ Include text labels for regions. Provide data table alternative. D3.js, Mapbox, Leaflet Pan + Zoom + Drill
8 7 Funnel/Flow funnel/flow Funnel Chart, Sankey Waterfall (for flows) Stages: gradient (starting color → ending color). Show conversion % ⚡ Good ✓ Clear stage labels + percentages. Good for accessibility if labeled. D3.js, Recharts, Custom SVG Hover + Drill
9 8 Performance vs Target performance-vs-target Gauge Chart or Bullet Chart Dial, Thermometer Performance: Red→Yellow→Green gradient. Target: marker line. Threshold colors ⚡ Good ✓ Add numerical value + percentage label beside gauge. D3.js, ApexCharts, Custom SVG Hover
10 9 Time-Series Forecast time-series-forecast Line with Confidence Band Ribbon Chart Actual: solid line #0080FF. Forecast: dashed #FF9500. Band: light shading ⚡ Good ✓ Clearly distinguish actual vs forecast. Add legend. Chart.js, ApexCharts, Plotly Hover + Toggle
11 10 Anomaly Detection anomaly-detection Line Chart with Highlights Scatter with Alert Normal: blue #0080FF. Anomaly: red #FF0000 circle/square marker + alert ⚡ Good ✓ Circle/marker for anomalies. Add text alert annotation. D3.js, Plotly, ApexCharts Hover + Alert
12 11 Hierarchical/Nested Data hierarchical/nested-data Treemap Sunburst, Nested Donut, Icicle Parent: distinct hues. Children: lighter shades. White borders 2-3px. ⚠ Moderate ⚠ Poor - provide table alternative. Label large areas. D3.js, Recharts, ApexCharts Hover + Drilldown
13 12 Flow/Process Data flow/process-data Sankey Diagram Alluvial, Chord Diagram Gradient from source to target. Opacity 0.4-0.6 for flows. ⚠ Moderate ⚠ Poor - provide flow table alternative. D3.js (d3-sankey), Plotly Hover + Drilldown
14 13 Cumulative Changes cumulative-changes Waterfall Chart Stacked Bar, Cascade Increases: #4CAF50. Decreases: #F44336. Start: #2196F3. End: #0D47A1. ⚡ Good ✓ Good - clear directional colors with labels. ApexCharts, Highcharts, Plotly Hover
15 14 Multi-Variable Comparison multi-variable-comparison Radar/Spider Chart Parallel Coordinates, Grouped Bar Single: #0080FF 20% fill. Multiple: distinct colors per dataset. ⚡ Good ⚠ Moderate - limit 5-8 axes. Add data table. Chart.js, Recharts, ApexCharts Hover + Toggle
16 15 Stock/Trading OHLC stock/trading-ohlc Candlestick Chart OHLC Bar, Heikin-Ashi Bullish: #26A69A. Bearish: #EF5350. Volume: 40% opacity below. ⚡ Good ⚠ Moderate - provide OHLC data table. Lightweight Charts (TradingView), ApexCharts Real-time + Hover + Zoom
17 16 Relationship/Connection Data relationship/connection-data Network Graph Hierarchical Tree, Adjacency Matrix Node types: categorical colors. Edges: #90A4AE 60% opacity. ❌ Poor (500+ nodes struggles) ❌ Very Poor - provide adjacency list alternative. D3.js (d3-force), Vis.js, Cytoscape.js Drilldown + Hover + Drag
18 17 Distribution/Statistical distribution/statistical Box Plot Violin Plot, Beeswarm Box: #BBDEFB. Border: #1976D2. Median: #D32F2F. Outliers: #F44336. ⚡ Excellent ✓ Good - include stats table (min, Q1, median, Q3, max). Plotly, D3.js, Chart.js (plugin) Hover
19 18 Performance vs Target (Compact) performance-vs-target-(compact) Bullet Chart Gauge, Progress Bar Ranges: #FFCDD2, #FFF9C4, #C8E6C9. Performance: #1976D2. Target: black 3px. ⚡ Excellent ✓ Excellent - compact with clear values. D3.js, Plotly, Custom SVG Hover
20 19 Proportional/Percentage proportional/percentage Waffle Chart Pictogram, Stacked Bar 100% 10x10 grid. 3-5 categories max. 2-3px spacing between squares. ⚡ Good ✓ Good - better than pie for accessibility. D3.js, React-Waffle, Custom CSS Grid Hover
21 20 Hierarchical Proportional hierarchical-proportional Sunburst Chart Treemap, Icicle, Circle Packing Center to outer: darker to lighter. 15-20% lighter per level. ⚠ Moderate ⚠ Poor - provide hierarchy table alternative. D3.js (d3-hierarchy), Recharts, ApexCharts Drilldown + Hover
22 21 Root Cause Analysis root cause, decomposition, tree, hierarchy, drill-down, ai-split Decomposition Tree Decision Tree, Flow Chart Nodes: #2563EB (Primary) vs #EF4444 (Negative impact). Connectors: Neutral grey. ⚠ Moderate (calculation heavy) ✓ clear hierarchy. Allow keyboard navigation for nodes. Power BI (native), React-Flow, Custom D3.js Drill + Expand
23 22 3D Spatial Data 3d, spatial, immersive, terrain, molecular, volumetric 3D Scatter/Surface Plot Volumetric Rendering, Point Cloud Depth cues: lighting/shading. Z-axis: color gradient (cool to warm). ❌ Heavy (WebGL required) ❌ Poor - requires alternative 2D view or data table. Three.js, Deck.gl, Plotly 3D Rotate + Zoom + VR
24 23 Real-Time Streaming streaming, real-time, ticker, live, velocity, pulse Streaming Area Chart Ticker Tape, Moving Gauge Current: Bright Pulse (#00FF00). History: Fading opacity. Grid: Dark. ⚡ Optimized (canvas/webgl) ⚠ Flashing elements - provide pause button. High contrast. Smoothed D3.js CanvasJS
25 24 Sentiment/Emotion sentiment, emotion, nlp, opinion, feeling Word Cloud with Sentiment Sentiment Arc, Radar Chart Positive: #22C55E. Negative: #EF4444. Neutral: #94A3B8. Size = Frequency. ⚡ Good ⚠ Word clouds poor for screen readers. Use list view. D3-cloud, Highcharts, Nivo Hover + Filter
26 25 Process Mining process, mining, variants, path, bottleneck, log Process Map / Graph Directed Acyclic Graph (DAG), Petri Net Happy path: #10B981 (Thick). Deviations: #F59E0B (Thin). Bottlenecks: #EF4444. ⚠ Moderate to Heavy ⚠ Complex graphs hard to navigate. Provide path summary. React-Flow, Cytoscape.js, Recharts Drag + Node-Click

97
skills/ui-ux-pro-max/data/colors.csv
View File

@ -1,97 +0,0 @@
No,Product Type,Primary (Hex),Secondary (Hex),CTA (Hex),Background (Hex),Text (Hex),Border (Hex),Notes
1,SaaS (General),#2563EB,#3B82F6,#F97316,#F8FAFC,#1E293B,#E2E8F0,Trust blue + orange CTA contrast
2,Micro SaaS,#6366F1,#818CF8,#10B981,#F5F3FF,#1E1B4B,#E0E7FF,Indigo primary + emerald CTA
3,E-commerce,#059669,#10B981,#F97316,#ECFDF5,#064E3B,#A7F3D0,Success green + urgency orange
4,E-commerce Luxury,#1C1917,#44403C,#CA8A04,#FAFAF9,#0C0A09,#D6D3D1,Premium dark + gold accent
5,Service Landing Page,#0EA5E9,#38BDF8,#F97316,#F0F9FF,#0C4A6E,#BAE6FD,Sky blue trust + warm CTA
6,B2B Service,#0F172A,#334155,#0369A1,#F8FAFC,#020617,#E2E8F0,Professional navy + blue CTA
7,Financial Dashboard,#0F172A,#1E293B,#22C55E,#020617,#F8FAFC,#334155,Dark bg + green positive indicators
8,Analytics Dashboard,#1E40AF,#3B82F6,#F59E0B,#F8FAFC,#1E3A8A,#DBEAFE,Blue data + amber highlights
9,Healthcare App,#0891B2,#22D3EE,#059669,#ECFEFF,#164E63,#A5F3FC,Calm cyan + health green
10,Educational App,#4F46E5,#818CF8,#F97316,#EEF2FF,#1E1B4B,#C7D2FE,Playful indigo + energetic orange
11,Creative Agency,#EC4899,#F472B6,#06B6D4,#FDF2F8,#831843,#FBCFE8,Bold pink + cyan accent
12,Portfolio/Personal,#18181B,#3F3F46,#2563EB,#FAFAFA,#09090B,#E4E4E7,Monochrome + blue accent
13,Gaming,#7C3AED,#A78BFA,#F43F5E,#0F0F23,#E2E8F0,#4C1D95,Neon purple + rose action
14,Government/Public Service,#0F172A,#334155,#0369A1,#F8FAFC,#020617,#E2E8F0,High contrast navy + blue
15,Fintech/Crypto,#F59E0B,#FBBF24,#8B5CF6,#0F172A,#F8FAFC,#334155,Gold trust + purple tech
16,Social Media App,#E11D48,#FB7185,#2563EB,#FFF1F2,#881337,#FECDD3,Vibrant rose + engagement blue
17,Productivity Tool,#0D9488,#14B8A6,#F97316,#F0FDFA,#134E4A,#99F6E4,Teal focus + action orange
18,Design System/Component Library,#4F46E5,#6366F1,#F97316,#EEF2FF,#312E81,#C7D2FE,Indigo brand + doc hierarchy
19,AI/Chatbot Platform,#7C3AED,#A78BFA,#06B6D4,#FAF5FF,#1E1B4B,#DDD6FE,AI purple + cyan interactions
20,NFT/Web3 Platform,#8B5CF6,#A78BFA,#FBBF24,#0F0F23,#F8FAFC,#4C1D95,Purple tech + gold value
21,Creator Economy Platform,#EC4899,#F472B6,#F97316,#FDF2F8,#831843,#FBCFE8,Creator pink + engagement orange
22,Sustainability/ESG Platform,#059669,#10B981,#0891B2,#ECFDF5,#064E3B,#A7F3D0,Nature green + ocean blue
23,Remote Work/Collaboration Tool,#6366F1,#818CF8,#10B981,#F5F3FF,#312E81,#E0E7FF,Calm indigo + success green
24,Mental Health App,#8B5CF6,#C4B5FD,#10B981,#FAF5FF,#4C1D95,#EDE9FE,Calming lavender + wellness green
25,Pet Tech App,#F97316,#FB923C,#2563EB,#FFF7ED,#9A3412,#FED7AA,Playful orange + trust blue
26,Smart Home/IoT Dashboard,#1E293B,#334155,#22C55E,#0F172A,#F8FAFC,#475569,Dark tech + status green
27,EV/Charging Ecosystem,#0891B2,#22D3EE,#22C55E,#ECFEFF,#164E63,#A5F3FC,Electric cyan + eco green
28,Subscription Box Service,#D946EF,#E879F9,#F97316,#FDF4FF,#86198F,#F5D0FE,Excitement purple + urgency orange
29,Podcast Platform,#1E1B4B,#312E81,#F97316,#0F0F23,#F8FAFC,#4338CA,Dark audio + warm accent
30,Dating App,#E11D48,#FB7185,#F97316,#FFF1F2,#881337,#FECDD3,Romantic rose + warm orange
31,Micro-Credentials/Badges Platform,#0369A1,#0EA5E9,#CA8A04,#F0F9FF,#0C4A6E,#BAE6FD,Trust blue + achievement gold
32,Knowledge Base/Documentation,#475569,#64748B,#2563EB,#F8FAFC,#1E293B,#E2E8F0,Neutral grey + link blue
33,Hyperlocal Services,#059669,#10B981,#F97316,#ECFDF5,#064E3B,#A7F3D0,Location green + action orange
34,Beauty/Spa/Wellness Service,#EC4899,#F9A8D4,#8B5CF6,#FDF2F8,#831843,#FBCFE8,Soft pink + lavender luxury
35,Luxury/Premium Brand,#1C1917,#44403C,#CA8A04,#FAFAF9,#0C0A09,#D6D3D1,Premium black + gold accent
36,Restaurant/Food Service,#DC2626,#F87171,#CA8A04,#FEF2F2,#450A0A,#FECACA,Appetizing red + warm gold
37,Fitness/Gym App,#F97316,#FB923C,#22C55E,#1F2937,#F8FAFC,#374151,Energy orange + success green
38,Real Estate/Property,#0F766E,#14B8A6,#0369A1,#F0FDFA,#134E4A,#99F6E4,Trust teal + professional blue
39,Travel/Tourism Agency,#0EA5E9,#38BDF8,#F97316,#F0F9FF,#0C4A6E,#BAE6FD,Sky blue + adventure orange
40,Hotel/Hospitality,#1E3A8A,#3B82F6,#CA8A04,#F8FAFC,#1E40AF,#BFDBFE,Luxury navy + gold service
41,Wedding/Event Planning,#DB2777,#F472B6,#CA8A04,#FDF2F8,#831843,#FBCFE8,Romantic pink + elegant gold
42,Legal Services,#1E3A8A,#1E40AF,#B45309,#F8FAFC,#0F172A,#CBD5E1,Authority navy + trust gold
43,Insurance Platform,#0369A1,#0EA5E9,#22C55E,#F0F9FF,#0C4A6E,#BAE6FD,Security blue + protected green
44,Banking/Traditional Finance,#0F172A,#1E3A8A,#CA8A04,#F8FAFC,#020617,#E2E8F0,Trust navy + premium gold
45,Online Course/E-learning,#0D9488,#2DD4BF,#F97316,#F0FDFA,#134E4A,#5EEAD4,Progress teal + achievement orange
46,Non-profit/Charity,#0891B2,#22D3EE,#F97316,#ECFEFF,#164E63,#A5F3FC,Compassion blue + action orange
47,Music Streaming,#1E1B4B,#4338CA,#22C55E,#0F0F23,#F8FAFC,#312E81,Dark audio + play green
48,Video Streaming/OTT,#0F0F23,#1E1B4B,#E11D48,#000000,#F8FAFC,#312E81,Cinema dark + play red
49,Job Board/Recruitment,#0369A1,#0EA5E9,#22C55E,#F0F9FF,#0C4A6E,#BAE6FD,Professional blue + success green
50,Marketplace (P2P),#7C3AED,#A78BFA,#22C55E,#FAF5FF,#4C1D95,#DDD6FE,Trust purple + transaction green
51,Logistics/Delivery,#2563EB,#3B82F6,#F97316,#EFF6FF,#1E40AF,#BFDBFE,Tracking blue + delivery orange
52,Agriculture/Farm Tech,#15803D,#22C55E,#CA8A04,#F0FDF4,#14532D,#BBF7D0,Earth green + harvest gold
53,Construction/Architecture,#64748B,#94A3B8,#F97316,#F8FAFC,#334155,#E2E8F0,Industrial grey + safety orange
54,Automotive/Car Dealership,#1E293B,#334155,#DC2626,#F8FAFC,#0F172A,#E2E8F0,Premium dark + action red
55,Photography Studio,#18181B,#27272A,#F8FAFC,#000000,#FAFAFA,#3F3F46,Pure black + white contrast
56,Coworking Space,#F59E0B,#FBBF24,#2563EB,#FFFBEB,#78350F,#FDE68A,Energetic amber + booking blue
57,Cleaning Service,#0891B2,#22D3EE,#22C55E,#ECFEFF,#164E63,#A5F3FC,Fresh cyan + clean green
58,Home Services (Plumber/Electrician),#1E40AF,#3B82F6,#F97316,#EFF6FF,#1E3A8A,#BFDBFE,Professional blue + urgent orange
59,Childcare/Daycare,#F472B6,#FBCFE8,#22C55E,#FDF2F8,#9D174D,#FCE7F3,Soft pink + safe green
60,Senior Care/Elderly,#0369A1,#38BDF8,#22C55E,#F0F9FF,#0C4A6E,#E0F2FE,Calm blue + reassuring green
61,Medical Clinic,#0891B2,#22D3EE,#22C55E,#F0FDFA,#134E4A,#CCFBF1,Medical teal + health green
62,Pharmacy/Drug Store,#15803D,#22C55E,#0369A1,#F0FDF4,#14532D,#BBF7D0,Pharmacy green + trust blue
63,Dental Practice,#0EA5E9,#38BDF8,#FBBF24,#F0F9FF,#0C4A6E,#BAE6FD,Fresh blue + smile yellow
64,Veterinary Clinic,#0D9488,#14B8A6,#F97316,#F0FDFA,#134E4A,#99F6E4,Caring teal + warm orange
65,Florist/Plant Shop,#15803D,#22C55E,#EC4899,#F0FDF4,#14532D,#BBF7D0,Natural green + floral pink
66,Bakery/Cafe,#92400E,#B45309,#F8FAFC,#FEF3C7,#78350F,#FDE68A,Warm brown + cream white
67,Coffee Shop,#78350F,#92400E,#FBBF24,#FEF3C7,#451A03,#FDE68A,Coffee brown + warm gold
68,Brewery/Winery,#7C2D12,#B91C1C,#CA8A04,#FEF2F2,#450A0A,#FECACA,Deep burgundy + craft gold
69,Airline,#1E3A8A,#3B82F6,#F97316,#EFF6FF,#1E40AF,#BFDBFE,Sky blue + booking orange
70,News/Media Platform,#DC2626,#EF4444,#1E40AF,#FEF2F2,#450A0A,#FECACA,Breaking red + link blue
71,Magazine/Blog,#18181B,#3F3F46,#EC4899,#FAFAFA,#09090B,#E4E4E7,Editorial black + accent pink
72,Freelancer Platform,#6366F1,#818CF8,#22C55E,#EEF2FF,#312E81,#C7D2FE,Creative indigo + hire green
73,Consulting Firm,#0F172A,#334155,#CA8A04,#F8FAFC,#020617,#E2E8F0,Authority navy + premium gold
74,Marketing Agency,#EC4899,#F472B6,#06B6D4,#FDF2F8,#831843,#FBCFE8,Bold pink + creative cyan
75,Event Management,#7C3AED,#A78BFA,#F97316,#FAF5FF,#4C1D95,#DDD6FE,Excitement purple + action orange
76,Conference/Webinar Platform,#1E40AF,#3B82F6,#22C55E,#EFF6FF,#1E3A8A,#BFDBFE,Professional blue + join green
77,Membership/Community,#7C3AED,#A78BFA,#22C55E,#FAF5FF,#4C1D95,#DDD6FE,Community purple + join green
78,Newsletter Platform,#0369A1,#0EA5E9,#F97316,#F0F9FF,#0C4A6E,#BAE6FD,Trust blue + subscribe orange
79,Digital Products/Downloads,#6366F1,#818CF8,#22C55E,#EEF2FF,#312E81,#C7D2FE,Digital indigo + buy green
80,Church/Religious Organization,#7C3AED,#A78BFA,#CA8A04,#FAF5FF,#4C1D95,#DDD6FE,Spiritual purple + warm gold
81,Sports Team/Club,#DC2626,#EF4444,#FBBF24,#FEF2F2,#7F1D1D,#FECACA,Team red + championship gold
82,Museum/Gallery,#18181B,#27272A,#F8FAFC,#FAFAFA,#09090B,#E4E4E7,Gallery black + white space
83,Theater/Cinema,#1E1B4B,#312E81,#CA8A04,#0F0F23,#F8FAFC,#4338CA,Dramatic dark + spotlight gold
84,Language Learning App,#4F46E5,#818CF8,#22C55E,#EEF2FF,#312E81,#C7D2FE,Learning indigo + progress green
85,Coding Bootcamp,#0F172A,#1E293B,#22C55E,#020617,#F8FAFC,#334155,Terminal dark + success green
86,Cybersecurity Platform,#00FF41,#0D0D0D,#FF3333,#000000,#E0E0E0,#1F1F1F,Matrix green + alert red
87,Developer Tool / IDE,#1E293B,#334155,#22C55E,#0F172A,#F8FAFC,#475569,Code dark + run green
88,Biotech / Life Sciences,#0EA5E9,#0284C7,#10B981,#F0F9FF,#0C4A6E,#BAE6FD,DNA blue + life green
89,Space Tech / Aerospace,#F8FAFC,#94A3B8,#3B82F6,#0B0B10,#F8FAFC,#1E293B,Star white + launch blue
90,Architecture / Interior,#171717,#404040,#D4AF37,#FFFFFF,#171717,#E5E5E5,Minimal black + accent gold
91,Quantum Computing,#00FFFF,#7B61FF,#FF00FF,#050510,#E0E0FF,#333344,Quantum cyan + interference purple
92,Biohacking / Longevity,#FF4D4D,#4D94FF,#00E676,#F5F5F7,#1C1C1E,#E5E5EA,Bio red/blue + vitality green
93,Autonomous Systems,#00FF41,#008F11,#FF3333,#0D1117,#E6EDF3,#30363D,Terminal green + alert red
94,Generative AI Art,#18181B,#3F3F46,#EC4899,#FAFAFA,#09090B,#E4E4E7,Canvas neutral + creative pink
95,Spatial / Vision OS,#FFFFFF,#E5E5E5,#007AFF,#888888,#000000,#CCCCCC,Glass white + system blue
96,Climate Tech,#059669,#10B981,#FBBF24,#ECFDF5,#064E3B,#A7F3D0,Nature green + solar gold
1 No Product Type Primary (Hex) Secondary (Hex) CTA (Hex) Background (Hex) Text (Hex) Border (Hex) Notes
2 1 SaaS (General) #2563EB #3B82F6 #F97316 #F8FAFC #1E293B #E2E8F0 Trust blue + orange CTA contrast
3 2 Micro SaaS #6366F1 #818CF8 #10B981 #F5F3FF #1E1B4B #E0E7FF Indigo primary + emerald CTA
4 3 E-commerce #059669 #10B981 #F97316 #ECFDF5 #064E3B #A7F3D0 Success green + urgency orange
5 4 E-commerce Luxury #1C1917 #44403C #CA8A04 #FAFAF9 #0C0A09 #D6D3D1 Premium dark + gold accent
6 5 Service Landing Page #0EA5E9 #38BDF8 #F97316 #F0F9FF #0C4A6E #BAE6FD Sky blue trust + warm CTA
7 6 B2B Service #0F172A #334155 #0369A1 #F8FAFC #020617 #E2E8F0 Professional navy + blue CTA
8 7 Financial Dashboard #0F172A #1E293B #22C55E #020617 #F8FAFC #334155 Dark bg + green positive indicators
9 8 Analytics Dashboard #1E40AF #3B82F6 #F59E0B #F8FAFC #1E3A8A #DBEAFE Blue data + amber highlights
10 9 Healthcare App #0891B2 #22D3EE #059669 #ECFEFF #164E63 #A5F3FC Calm cyan + health green
11 10 Educational App #4F46E5 #818CF8 #F97316 #EEF2FF #1E1B4B #C7D2FE Playful indigo + energetic orange
12 11 Creative Agency #EC4899 #F472B6 #06B6D4 #FDF2F8 #831843 #FBCFE8 Bold pink + cyan accent
13 12 Portfolio/Personal #18181B #3F3F46 #2563EB #FAFAFA #09090B #E4E4E7 Monochrome + blue accent
14 13 Gaming #7C3AED #A78BFA #F43F5E #0F0F23 #E2E8F0 #4C1D95 Neon purple + rose action
15 14 Government/Public Service #0F172A #334155 #0369A1 #F8FAFC #020617 #E2E8F0 High contrast navy + blue
16 15 Fintech/Crypto #F59E0B #FBBF24 #8B5CF6 #0F172A #F8FAFC #334155 Gold trust + purple tech
17 16 Social Media App #E11D48 #FB7185 #2563EB #FFF1F2 #881337 #FECDD3 Vibrant rose + engagement blue
18 17 Productivity Tool #0D9488 #14B8A6 #F97316 #F0FDFA #134E4A #99F6E4 Teal focus + action orange
19 18 Design System/Component Library #4F46E5 #6366F1 #F97316 #EEF2FF #312E81 #C7D2FE Indigo brand + doc hierarchy
20 19 AI/Chatbot Platform #7C3AED #A78BFA #06B6D4 #FAF5FF #1E1B4B #DDD6FE AI purple + cyan interactions
21 20 NFT/Web3 Platform #8B5CF6 #A78BFA #FBBF24 #0F0F23 #F8FAFC #4C1D95 Purple tech + gold value
22 21 Creator Economy Platform #EC4899 #F472B6 #F97316 #FDF2F8 #831843 #FBCFE8 Creator pink + engagement orange
23 22 Sustainability/ESG Platform #059669 #10B981 #0891B2 #ECFDF5 #064E3B #A7F3D0 Nature green + ocean blue
24 23 Remote Work/Collaboration Tool #6366F1 #818CF8 #10B981 #F5F3FF #312E81 #E0E7FF Calm indigo + success green
25 24 Mental Health App #8B5CF6 #C4B5FD #10B981 #FAF5FF #4C1D95 #EDE9FE Calming lavender + wellness green
26 25 Pet Tech App #F97316 #FB923C #2563EB #FFF7ED #9A3412 #FED7AA Playful orange + trust blue
27 26 Smart Home/IoT Dashboard #1E293B #334155 #22C55E #0F172A #F8FAFC #475569 Dark tech + status green
28 27 EV/Charging Ecosystem #0891B2 #22D3EE #22C55E #ECFEFF #164E63 #A5F3FC Electric cyan + eco green
29 28 Subscription Box Service #D946EF #E879F9 #F97316 #FDF4FF #86198F #F5D0FE Excitement purple + urgency orange
30 29 Podcast Platform #1E1B4B #312E81 #F97316 #0F0F23 #F8FAFC #4338CA Dark audio + warm accent
31 30 Dating App #E11D48 #FB7185 #F97316 #FFF1F2 #881337 #FECDD3 Romantic rose + warm orange
32 31 Micro-Credentials/Badges Platform #0369A1 #0EA5E9 #CA8A04 #F0F9FF #0C4A6E #BAE6FD Trust blue + achievement gold
33 32 Knowledge Base/Documentation #475569 #64748B #2563EB #F8FAFC #1E293B #E2E8F0 Neutral grey + link blue
34 33 Hyperlocal Services #059669 #10B981 #F97316 #ECFDF5 #064E3B #A7F3D0 Location green + action orange
35 34 Beauty/Spa/Wellness Service #EC4899 #F9A8D4 #8B5CF6 #FDF2F8 #831843 #FBCFE8 Soft pink + lavender luxury
36 35 Luxury/Premium Brand #1C1917 #44403C #CA8A04 #FAFAF9 #0C0A09 #D6D3D1 Premium black + gold accent
37 36 Restaurant/Food Service #DC2626 #F87171 #CA8A04 #FEF2F2 #450A0A #FECACA Appetizing red + warm gold
38 37 Fitness/Gym App #F97316 #FB923C #22C55E #1F2937 #F8FAFC #374151 Energy orange + success green
39 38 Real Estate/Property #0F766E #14B8A6 #0369A1 #F0FDFA #134E4A #99F6E4 Trust teal + professional blue
40 39 Travel/Tourism Agency #0EA5E9 #38BDF8 #F97316 #F0F9FF #0C4A6E #BAE6FD Sky blue + adventure orange
41 40 Hotel/Hospitality #1E3A8A #3B82F6 #CA8A04 #F8FAFC #1E40AF #BFDBFE Luxury navy + gold service
42 41 Wedding/Event Planning #DB2777 #F472B6 #CA8A04 #FDF2F8 #831843 #FBCFE8 Romantic pink + elegant gold
43 42 Legal Services #1E3A8A #1E40AF #B45309 #F8FAFC #0F172A #CBD5E1 Authority navy + trust gold
44 43 Insurance Platform #0369A1 #0EA5E9 #22C55E #F0F9FF #0C4A6E #BAE6FD Security blue + protected green
45 44 Banking/Traditional Finance #0F172A #1E3A8A #CA8A04 #F8FAFC #020617 #E2E8F0 Trust navy + premium gold
46 45 Online Course/E-learning #0D9488 #2DD4BF #F97316 #F0FDFA #134E4A #5EEAD4 Progress teal + achievement orange
47 46 Non-profit/Charity #0891B2 #22D3EE #F97316 #ECFEFF #164E63 #A5F3FC Compassion blue + action orange
48 47 Music Streaming #1E1B4B #4338CA #22C55E #0F0F23 #F8FAFC #312E81 Dark audio + play green
49 48 Video Streaming/OTT #0F0F23 #1E1B4B #E11D48 #000000 #F8FAFC #312E81 Cinema dark + play red
50 49 Job Board/Recruitment #0369A1 #0EA5E9 #22C55E #F0F9FF #0C4A6E #BAE6FD Professional blue + success green
51 50 Marketplace (P2P) #7C3AED #A78BFA #22C55E #FAF5FF #4C1D95 #DDD6FE Trust purple + transaction green
52 51 Logistics/Delivery #2563EB #3B82F6 #F97316 #EFF6FF #1E40AF #BFDBFE Tracking blue + delivery orange
53 52 Agriculture/Farm Tech #15803D #22C55E #CA8A04 #F0FDF4 #14532D #BBF7D0 Earth green + harvest gold
54 53 Construction/Architecture #64748B #94A3B8 #F97316 #F8FAFC #334155 #E2E8F0 Industrial grey + safety orange
55 54 Automotive/Car Dealership #1E293B #334155 #DC2626 #F8FAFC #0F172A #E2E8F0 Premium dark + action red
56 55 Photography Studio #18181B #27272A #F8FAFC #000000 #FAFAFA #3F3F46 Pure black + white contrast
57 56 Coworking Space #F59E0B #FBBF24 #2563EB #FFFBEB #78350F #FDE68A Energetic amber + booking blue
58 57 Cleaning Service #0891B2 #22D3EE #22C55E #ECFEFF #164E63 #A5F3FC Fresh cyan + clean green
59 58 Home Services (Plumber/Electrician) #1E40AF #3B82F6 #F97316 #EFF6FF #1E3A8A #BFDBFE Professional blue + urgent orange
60 59 Childcare/Daycare #F472B6 #FBCFE8 #22C55E #FDF2F8 #9D174D #FCE7F3 Soft pink + safe green
61 60 Senior Care/Elderly #0369A1 #38BDF8 #22C55E #F0F9FF #0C4A6E #E0F2FE Calm blue + reassuring green
62 61 Medical Clinic #0891B2 #22D3EE #22C55E #F0FDFA #134E4A #CCFBF1 Medical teal + health green
63 62 Pharmacy/Drug Store #15803D #22C55E #0369A1 #F0FDF4 #14532D #BBF7D0 Pharmacy green + trust blue
64 63 Dental Practice #0EA5E9 #38BDF8 #FBBF24 #F0F9FF #0C4A6E #BAE6FD Fresh blue + smile yellow
65 64 Veterinary Clinic #0D9488 #14B8A6 #F97316 #F0FDFA #134E4A #99F6E4 Caring teal + warm orange
66 65 Florist/Plant Shop #15803D #22C55E #EC4899 #F0FDF4 #14532D #BBF7D0 Natural green + floral pink
67 66 Bakery/Cafe #92400E #B45309 #F8FAFC #FEF3C7 #78350F #FDE68A Warm brown + cream white
68 67 Coffee Shop #78350F #92400E #FBBF24 #FEF3C7 #451A03 #FDE68A Coffee brown + warm gold
69 68 Brewery/Winery #7C2D12 #B91C1C #CA8A04 #FEF2F2 #450A0A #FECACA Deep burgundy + craft gold
70 69 Airline #1E3A8A #3B82F6 #F97316 #EFF6FF #1E40AF #BFDBFE Sky blue + booking orange
71 70 News/Media Platform #DC2626 #EF4444 #1E40AF #FEF2F2 #450A0A #FECACA Breaking red + link blue
72 71 Magazine/Blog #18181B #3F3F46 #EC4899 #FAFAFA #09090B #E4E4E7 Editorial black + accent pink
73 72 Freelancer Platform #6366F1 #818CF8 #22C55E #EEF2FF #312E81 #C7D2FE Creative indigo + hire green
74 73 Consulting Firm #0F172A #334155 #CA8A04 #F8FAFC #020617 #E2E8F0 Authority navy + premium gold
75 74 Marketing Agency #EC4899 #F472B6 #06B6D4 #FDF2F8 #831843 #FBCFE8 Bold pink + creative cyan
76 75 Event Management #7C3AED #A78BFA #F97316 #FAF5FF #4C1D95 #DDD6FE Excitement purple + action orange
77 76 Conference/Webinar Platform #1E40AF #3B82F6 #22C55E #EFF6FF #1E3A8A #BFDBFE Professional blue + join green
78 77 Membership/Community #7C3AED #A78BFA #22C55E #FAF5FF #4C1D95 #DDD6FE Community purple + join green
79 78 Newsletter Platform #0369A1 #0EA5E9 #F97316 #F0F9FF #0C4A6E #BAE6FD Trust blue + subscribe orange
80 79 Digital Products/Downloads #6366F1 #818CF8 #22C55E #EEF2FF #312E81 #C7D2FE Digital indigo + buy green
81 80 Church/Religious Organization #7C3AED #A78BFA #CA8A04 #FAF5FF #4C1D95 #DDD6FE Spiritual purple + warm gold
82 81 Sports Team/Club #DC2626 #EF4444 #FBBF24 #FEF2F2 #7F1D1D #FECACA Team red + championship gold
83 82 Museum/Gallery #18181B #27272A #F8FAFC #FAFAFA #09090B #E4E4E7 Gallery black + white space
84 83 Theater/Cinema #1E1B4B #312E81 #CA8A04 #0F0F23 #F8FAFC #4338CA Dramatic dark + spotlight gold
85 84 Language Learning App #4F46E5 #818CF8 #22C55E #EEF2FF #312E81 #C7D2FE Learning indigo + progress green
86 85 Coding Bootcamp #0F172A #1E293B #22C55E #020617 #F8FAFC #334155 Terminal dark + success green
87 86 Cybersecurity Platform #00FF41 #0D0D0D #FF3333 #000000 #E0E0E0 #1F1F1F Matrix green + alert red
88 87 Developer Tool / IDE #1E293B #334155 #22C55E #0F172A #F8FAFC #475569 Code dark + run green
89 88 Biotech / Life Sciences #0EA5E9 #0284C7 #10B981 #F0F9FF #0C4A6E #BAE6FD DNA blue + life green
90 89 Space Tech / Aerospace #F8FAFC #94A3B8 #3B82F6 #0B0B10 #F8FAFC #1E293B Star white + launch blue
91 90 Architecture / Interior #171717 #404040 #D4AF37 #FFFFFF #171717 #E5E5E5 Minimal black + accent gold
92 91 Quantum Computing #00FFFF #7B61FF #FF00FF #050510 #E0E0FF #333344 Quantum cyan + interference purple
93 92 Biohacking / Longevity #FF4D4D #4D94FF #00E676 #F5F5F7 #1C1C1E #E5E5EA Bio red/blue + vitality green
94 93 Autonomous Systems #00FF41 #008F11 #FF3333 #0D1117 #E6EDF3 #30363D Terminal green + alert red
95 94 Generative AI Art #18181B #3F3F46 #EC4899 #FAFAFA #09090B #E4E4E7 Canvas neutral + creative pink
96 95 Spatial / Vision OS #FFFFFF #E5E5E5 #007AFF #888888 #000000 #CCCCCC Glass white + system blue
97 96 Climate Tech #059669 #10B981 #FBBF24 #ECFDF5 #064E3B #A7F3D0 Nature green + solar gold

101
skills/ui-ux-pro-max/data/icons.csv
View File

@ -1,101 +0,0 @@
No,Category,Icon Name,Keywords,Library,Import Code,Usage,Best For,Style
1,Navigation,menu,hamburger menu navigation toggle bars,Lucide,import { Menu } from 'lucide-react',<Menu />,Mobile navigation drawer toggle sidebar,Outline
2,Navigation,arrow-left,back previous return navigate,Lucide,import { ArrowLeft } from 'lucide-react',<ArrowLeft />,Back button breadcrumb navigation,Outline
3,Navigation,arrow-right,next forward continue navigate,Lucide,import { ArrowRight } from 'lucide-react',<ArrowRight />,Forward button next step CTA,Outline
4,Navigation,chevron-down,dropdown expand accordion select,Lucide,import { ChevronDown } from 'lucide-react',<ChevronDown />,Dropdown toggle accordion header,Outline
5,Navigation,chevron-up,collapse close accordion minimize,Lucide,import { ChevronUp } from 'lucide-react',<ChevronUp />,Accordion collapse minimize,Outline
6,Navigation,home,homepage main dashboard start,Lucide,import { Home } from 'lucide-react',<Home />,Home navigation main page,Outline
7,Navigation,x,close cancel dismiss remove exit,Lucide,import { X } from 'lucide-react',<X />,Modal close dismiss button,Outline
8,Navigation,external-link,open new tab external link,Lucide,import { ExternalLink } from 'lucide-react',<ExternalLink />,External link indicator,Outline
9,Action,plus,add create new insert,Lucide,import { Plus } from 'lucide-react',<Plus />,Add button create new item,Outline
10,Action,minus,remove subtract decrease delete,Lucide,import { Minus } from 'lucide-react',<Minus />,Remove item quantity decrease,Outline
11,Action,trash-2,delete remove discard bin,Lucide,import { Trash2 } from 'lucide-react',<Trash2 />,Delete action destructive,Outline
12,Action,edit,pencil modify change update,Lucide,import { Edit } from 'lucide-react',<Edit />,Edit button modify content,Outline
13,Action,save,disk store persist save,Lucide,import { Save } from 'lucide-react',<Save />,Save button persist changes,Outline
14,Action,download,export save file download,Lucide,import { Download } from 'lucide-react',<Download />,Download file export,Outline
15,Action,upload,import file attach upload,Lucide,import { Upload } from 'lucide-react',<Upload />,Upload file import,Outline
16,Action,copy,duplicate clipboard paste,Lucide,import { Copy } from 'lucide-react',<Copy />,Copy to clipboard,Outline
17,Action,share,social distribute send,Lucide,import { Share } from 'lucide-react',<Share />,Share button social,Outline
18,Action,search,find lookup filter query,Lucide,import { Search } from 'lucide-react',<Search />,Search input bar,Outline
19,Action,filter,sort refine narrow options,Lucide,import { Filter } from 'lucide-react',<Filter />,Filter dropdown sort,Outline
20,Action,settings,gear cog preferences config,Lucide,import { Settings } from 'lucide-react',<Settings />,Settings page configuration,Outline
21,Status,check,success done complete verified,Lucide,import { Check } from 'lucide-react',<Check />,Success state checkmark,Outline
22,Status,check-circle,success verified approved complete,Lucide,import { CheckCircle } from 'lucide-react',<CheckCircle />,Success badge verified,Outline
23,Status,x-circle,error failed cancel rejected,Lucide,import { XCircle } from 'lucide-react',<XCircle />,Error state failed,Outline
24,Status,alert-triangle,warning caution attention danger,Lucide,import { AlertTriangle } from 'lucide-react',<AlertTriangle />,Warning message caution,Outline
25,Status,alert-circle,info notice information help,Lucide,import { AlertCircle } from 'lucide-react',<AlertCircle />,Info notice alert,Outline
26,Status,info,information help tooltip details,Lucide,import { Info } from 'lucide-react',<Info />,Information tooltip help,Outline
27,Status,loader,loading spinner processing wait,Lucide,import { Loader } from 'lucide-react',<Loader className="animate-spin" />,Loading state spinner,Outline
28,Status,clock,time schedule pending wait,Lucide,import { Clock } from 'lucide-react',<Clock />,Pending time schedule,Outline
29,Communication,mail,email message inbox letter,Lucide,import { Mail } from 'lucide-react',<Mail />,Email contact inbox,Outline
30,Communication,message-circle,chat comment bubble conversation,Lucide,import { MessageCircle } from 'lucide-react',<MessageCircle />,Chat comment message,Outline
31,Communication,phone,call mobile telephone contact,Lucide,import { Phone } from 'lucide-react',<Phone />,Phone contact call,Outline
32,Communication,send,submit dispatch message airplane,Lucide,import { Send } from 'lucide-react',<Send />,Send message submit,Outline
33,Communication,bell,notification alert ring reminder,Lucide,import { Bell } from 'lucide-react',<Bell />,Notification bell alert,Outline
34,User,user,profile account person avatar,Lucide,import { User } from 'lucide-react',<User />,User profile account,Outline
35,User,users,team group people members,Lucide,import { Users } from 'lucide-react',<Users />,Team group members,Outline
36,User,user-plus,add invite new member,Lucide,import { UserPlus } from 'lucide-react',<UserPlus />,Add user invite,Outline
37,User,log-in,signin authenticate enter,Lucide,import { LogIn } from 'lucide-react',<LogIn />,Login signin,Outline
38,User,log-out,signout exit leave logout,Lucide,import { LogOut } from 'lucide-react',<LogOut />,Logout signout,Outline
39,Media,image,photo picture gallery thumbnail,Lucide,import { Image } from 'lucide-react',<Image />,Image photo gallery,Outline
40,Media,video,movie film play record,Lucide,import { Video } from 'lucide-react',<Video />,Video player media,Outline
41,Media,play,start video audio media,Lucide,import { Play } from 'lucide-react',<Play />,Play button video audio,Outline
42,Media,pause,stop halt video audio,Lucide,import { Pause } from 'lucide-react',<Pause />,Pause button media,Outline
43,Media,volume-2,sound audio speaker music,Lucide,import { Volume2 } from 'lucide-react',<Volume2 />,Volume audio sound,Outline
44,Media,mic,microphone record voice audio,Lucide,import { Mic } from 'lucide-react',<Mic />,Microphone voice record,Outline
45,Media,camera,photo capture snapshot picture,Lucide,import { Camera } from 'lucide-react',<Camera />,Camera photo capture,Outline
46,Commerce,shopping-cart,cart checkout basket buy,Lucide,import { ShoppingCart } from 'lucide-react',<ShoppingCart />,Shopping cart e-commerce,Outline
47,Commerce,shopping-bag,purchase buy store bag,Lucide,import { ShoppingBag } from 'lucide-react',<ShoppingBag />,Shopping bag purchase,Outline
48,Commerce,credit-card,payment card checkout stripe,Lucide,import { CreditCard } from 'lucide-react',<CreditCard />,Payment credit card,Outline
49,Commerce,dollar-sign,money price currency cost,Lucide,import { DollarSign } from 'lucide-react',<DollarSign />,Price money currency,Outline
50,Commerce,tag,label price discount sale,Lucide,import { Tag } from 'lucide-react',<Tag />,Price tag label,Outline
51,Commerce,gift,present reward bonus offer,Lucide,import { Gift } from 'lucide-react',<Gift />,Gift reward offer,Outline
52,Commerce,percent,discount sale offer promo,Lucide,import { Percent } from 'lucide-react',<Percent />,Discount percentage sale,Outline
53,Data,bar-chart,analytics statistics graph metrics,Lucide,import { BarChart } from 'lucide-react',<BarChart />,Bar chart analytics,Outline
54,Data,pie-chart,statistics distribution breakdown,Lucide,import { PieChart } from 'lucide-react',<PieChart />,Pie chart distribution,Outline
55,Data,trending-up,growth increase positive trend,Lucide,import { TrendingUp } from 'lucide-react',<TrendingUp />,Growth trend positive,Outline
56,Data,trending-down,decline decrease negative trend,Lucide,import { TrendingDown } from 'lucide-react',<TrendingDown />,Decline trend negative,Outline
57,Data,activity,pulse heartbeat monitor live,Lucide,import { Activity } from 'lucide-react',<Activity />,Activity monitor pulse,Outline
58,Data,database,storage server data backend,Lucide,import { Database } from 'lucide-react',<Database />,Database storage,Outline
59,Files,file,document page paper doc,Lucide,import { File } from 'lucide-react',<File />,File document,Outline
60,Files,file-text,document text page article,Lucide,import { FileText } from 'lucide-react',<FileText />,Text document article,Outline
61,Files,folder,directory organize group files,Lucide,import { Folder } from 'lucide-react',<Folder />,Folder directory,Outline
62,Files,folder-open,expanded browse files view,Lucide,import { FolderOpen } from 'lucide-react',<FolderOpen />,Open folder browse,Outline
63,Files,paperclip,attachment attach file link,Lucide,import { Paperclip } from 'lucide-react',<Paperclip />,Attachment paperclip,Outline
64,Files,link,url hyperlink chain connect,Lucide,import { Link } from 'lucide-react',<Link />,Link URL hyperlink,Outline
65,Files,clipboard,paste copy buffer notes,Lucide,import { Clipboard } from 'lucide-react',<Clipboard />,Clipboard paste,Outline
66,Layout,grid,tiles gallery layout dashboard,Lucide,import { Grid } from 'lucide-react',<Grid />,Grid layout gallery,Outline
67,Layout,list,rows table lines items,Lucide,import { List } from 'lucide-react',<List />,List view rows,Outline
68,Layout,columns,layout split dual sidebar,Lucide,import { Columns } from 'lucide-react',<Columns />,Column layout split,Outline
69,Layout,maximize,fullscreen expand enlarge zoom,Lucide,import { Maximize } from 'lucide-react',<Maximize />,Fullscreen maximize,Outline
70,Layout,minimize,reduce shrink collapse exit,Lucide,import { Minimize } from 'lucide-react',<Minimize />,Minimize reduce,Outline
71,Layout,sidebar,panel drawer navigation menu,Lucide,import { Sidebar } from 'lucide-react',<Sidebar />,Sidebar panel,Outline
72,Social,heart,like love favorite wishlist,Lucide,import { Heart } from 'lucide-react',<Heart />,Like favorite love,Outline
73,Social,star,rating review favorite bookmark,Lucide,import { Star } from 'lucide-react',<Star />,Star rating favorite,Outline
74,Social,thumbs-up,like approve agree positive,Lucide,import { ThumbsUp } from 'lucide-react',<ThumbsUp />,Like approve thumb,Outline
75,Social,thumbs-down,dislike disapprove disagree negative,Lucide,import { ThumbsDown } from 'lucide-react',<ThumbsDown />,Dislike disapprove,Outline
76,Social,bookmark,save later favorite mark,Lucide,import { Bookmark } from 'lucide-react',<Bookmark />,Bookmark save,Outline
77,Social,flag,report mark important highlight,Lucide,import { Flag } from 'lucide-react',<Flag />,Flag report,Outline
78,Device,smartphone,mobile phone device touch,Lucide,import { Smartphone } from 'lucide-react',<Smartphone />,Mobile smartphone,Outline
79,Device,tablet,ipad device touch screen,Lucide,import { Tablet } from 'lucide-react',<Tablet />,Tablet device,Outline
80,Device,monitor,desktop screen computer display,Lucide,import { Monitor } from 'lucide-react',<Monitor />,Desktop monitor,Outline
81,Device,laptop,notebook computer portable device,Lucide,import { Laptop } from 'lucide-react',<Laptop />,Laptop computer,Outline
82,Device,printer,print document output paper,Lucide,import { Printer } from 'lucide-react',<Printer />,Printer print,Outline
83,Security,lock,secure password protected private,Lucide,import { Lock } from 'lucide-react',<Lock />,Lock secure,Outline
84,Security,unlock,open access unsecure public,Lucide,import { Unlock } from 'lucide-react',<Unlock />,Unlock open,Outline
85,Security,shield,protection security safe guard,Lucide,import { Shield } from 'lucide-react',<Shield />,Shield protection,Outline
86,Security,key,password access unlock login,Lucide,import { Key } from 'lucide-react',<Key />,Key password,Outline
87,Security,eye,view show visible password,Lucide,import { Eye } from 'lucide-react',<Eye />,Show password view,Outline
88,Security,eye-off,hide invisible password hidden,Lucide,import { EyeOff } from 'lucide-react',<EyeOff />,Hide password,Outline
89,Location,map-pin,location marker place address,Lucide,import { MapPin } from 'lucide-react',<MapPin />,Location pin marker,Outline
90,Location,map,directions navigate geography location,Lucide,import { Map } from 'lucide-react',<Map />,Map directions,Outline
91,Location,navigation,compass direction pointer arrow,Lucide,import { Navigation } from 'lucide-react',<Navigation />,Navigation compass,Outline
92,Location,globe,world international global web,Lucide,import { Globe } from 'lucide-react',<Globe />,Globe world,Outline
93,Time,calendar,date schedule event appointment,Lucide,import { Calendar } from 'lucide-react',<Calendar />,Calendar date,Outline
94,Time,refresh-cw,reload sync update refresh,Lucide,import { RefreshCw } from 'lucide-react',<RefreshCw />,Refresh reload,Outline
95,Time,rotate-ccw,undo back revert history,Lucide,import { RotateCcw } from 'lucide-react',<RotateCcw />,Undo revert,Outline
96,Time,rotate-cw,redo forward repeat history,Lucide,import { RotateCw } from 'lucide-react',<RotateCw />,Redo forward,Outline
97,Development,code,develop programming syntax html,Lucide,import { Code } from 'lucide-react',<Code />,Code development,Outline
98,Development,terminal,console cli command shell,Lucide,import { Terminal } from 'lucide-react',<Terminal />,Terminal console,Outline
99,Development,git-branch,version control branch merge,Lucide,import { GitBranch } from 'lucide-react',<GitBranch />,Git branch,Outline
100,Development,github,repository code open source,Lucide,import { Github } from 'lucide-react',<Github />,GitHub repository,Outline
Can't render this file because it contains an unexpected character in line 28 and column 113.

31
skills/ui-ux-pro-max/data/landing.csv
View File

@ -1,31 +0,0 @@
No,Pattern Name,Keywords,Section Order,Primary CTA Placement,Color Strategy,Recommended Effects,Conversion Optimization
1,Hero + Features + CTA,"hero, hero-centric, features, feature-rich, cta, call-to-action","1. Hero with headline/image, 2. Value prop, 3. Key features (3-5), 4. CTA section, 5. Footer",Hero (sticky) + Bottom,Hero: Brand primary or vibrant. Features: Card bg #FAFAFA. CTA: Contrasting accent color,"Hero parallax, feature card hover lift, CTA glow on hover",Deep CTA placement. Use contrasting color (at least 7:1 contrast ratio). Sticky navbar CTA.
2,Hero + Testimonials + CTA,"hero, testimonials, social-proof, trust, reviews, cta","1. Hero, 2. Problem statement, 3. Solution overview, 4. Testimonials carousel, 5. CTA",Hero (sticky) + Post-testimonials,"Hero: Brand color. Testimonials: Light bg #F5F5F5. Quotes: Italic, muted color #666. CTA: Vibrant","Testimonial carousel slide animations, quote marks animations, avatar fade-in",Social proof before CTA. Use 3-5 testimonials. Include photo + name + role. CTA after social proof.
3,Product Demo + Features,"demo, product-demo, features, showcase, interactive","1. Hero, 2. Product video/mockup (center), 3. Feature breakdown per section, 4. Comparison (optional), 5. CTA",Video center + CTA right/bottom,Video surround: Brand color overlay. Features: Icon color #0080FF. Text: Dark #222,"Video play button pulse, feature scroll reveals, demo interaction highlights",Embedded product demo increases engagement. Use interactive mockup if possible. Auto-play video muted.
4,Minimal Single Column,"minimal, simple, direct, single-column, clean","1. Hero headline, 2. Short description, 3. Benefit bullets (3 max), 4. CTA, 5. Footer","Center, large CTA button",Minimalist: Brand + white #FFFFFF + accent. Buttons: High contrast 7:1+. Text: Black/Dark grey,Minimal hover effects. Smooth scroll. CTA scale on hover (subtle),Single CTA focus. Large typography. Lots of whitespace. No nav clutter. Mobile-first.
5,Funnel (3-Step Conversion),"funnel, conversion, steps, wizard, onboarding","1. Hero, 2. Step 1 (problem), 3. Step 2 (solution), 4. Step 3 (action), 5. CTA progression",Each step: mini-CTA. Final: main CTA,"Step colors: 1 (Red/Problem), 2 (Orange/Process), 3 (Green/Solution). CTA: Brand color","Step number animations, progress bar fill, step transitions smooth scroll",Progressive disclosure. Show only essential info per step. Use progress indicators. Multiple CTAs.
6,Comparison Table + CTA,"comparison, table, compare, versus, cta","1. Hero, 2. Problem intro, 3. Comparison table (product vs competitors), 4. Pricing (optional), 5. CTA",Table: Right column. CTA: Below table,Table: Alternating rows (white/light grey). Your product: Highlight #FFFACD (light yellow) or green. Text: Dark,"Table row hover highlight, price toggle animations, feature checkmark animations",Use comparison to show unique value. Highlight your product row. Include 'free trial' in pricing row.
7,Lead Magnet + Form,"lead, form, signup, capture, email, magnet","1. Hero (benefit headline), 2. Lead magnet preview (ebook cover, checklist, etc), 3. Form (minimal fields), 4. CTA submit",Form CTA: Submit button,Lead magnet: Professional design. Form: Clean white bg. Inputs: Light border #CCCCCC. CTA: Brand color,"Form focus state animations, input validation animations, success confirmation animation",Form fields ≤ 3 for best conversion. Offer valuable lead magnet preview. Show form submission progress.
8,Pricing Page + CTA,"pricing, plans, tiers, comparison, cta","1. Hero (pricing headline), 2. Price comparison cards, 3. Feature comparison table, 4. FAQ section, 5. Final CTA",Each card: CTA button. Sticky CTA in nav,"Free: Grey, Starter: Blue, Pro: Green/Gold, Enterprise: Dark. Cards: 1px border, shadow","Price toggle animation (monthly/yearly), card comparison highlight, FAQ accordion open/close",Recommend starter plan (pre-select/highlight). Show annual discount (20-30%). Use FAQs to address concerns.
9,Video-First Hero,"video, hero, media, visual, engaging","1. Hero with video background, 2. Key features overlay, 3. Benefits section, 4. CTA",Overlay on video (center/bottom) + Bottom section,Dark overlay 60% on video. Brand accent for CTA. White text on dark.,"Video autoplay muted, parallax scroll, text fade-in on scroll",86% higher engagement with video. Add captions for accessibility. Compress video for performance.
10,Scroll-Triggered Storytelling,"storytelling, scroll, narrative, story, immersive","1. Intro hook, 2. Chapter 1 (problem), 3. Chapter 2 (journey), 4. Chapter 3 (solution), 5. Climax CTA",End of each chapter (mini) + Final climax CTA,Progressive reveal. Each chapter has distinct color. Building intensity.,"ScrollTrigger animations, parallax layers, progressive disclosure, chapter transitions",Narrative increases time-on-page 3x. Use progress indicator. Mobile: simplify animations.
11,AI Personalization Landing,"ai, personalization, smart, recommendation, dynamic","1. Dynamic hero (personalized), 2. Relevant features, 3. Tailored testimonials, 4. Smart CTA",Context-aware placement based on user segment,Adaptive based on user data. A/B test color variations per segment.,"Dynamic content swap, fade transitions, personalized product recommendations",20%+ conversion with personalization. Requires analytics integration. Fallback for new users.
12,Waitlist/Coming Soon,"waitlist, coming-soon, launch, early-access, notify","1. Hero with countdown, 2. Product teaser/preview, 3. Email capture form, 4. Social proof (waitlist count)",Email form prominent (above fold) + Sticky form on scroll,Anticipation: Dark + accent highlights. Countdown in brand color. Urgency indicators.,"Countdown timer animation, email validation feedback, success confetti, social share buttons",Scarcity + exclusivity. Show waitlist count. Early access benefits. Referral program.
13,Comparison Table Focus,"comparison, table, versus, compare, features","1. Hero (problem statement), 2. Comparison matrix (you vs competitors), 3. Feature deep-dive, 4. Winner CTA",After comparison table (highlighted row) + Bottom,Your product column highlighted (accent bg or green). Competitors neutral. Checkmarks green.,"Table row hover highlight, feature checkmark animations, sticky comparison header",Show value vs competitors. 35% higher conversion. Be factual. Include pricing if favorable.
14,Pricing-Focused Landing,"pricing, price, cost, plans, subscription","1. Hero (value proposition), 2. Pricing cards (3 tiers), 3. Feature comparison, 4. FAQ, 5. Final CTA",Each pricing card + Sticky CTA in nav + Bottom,Popular plan highlighted (brand color border/bg). Free: grey. Enterprise: dark/premium.,"Price toggle monthly/annual animation, card hover lift, FAQ accordion smooth open",Annual discount 20-30%. Recommend mid-tier (most popular badge). Address objections in FAQ.
15,App Store Style Landing,"app, mobile, download, store, install","1. Hero with device mockup, 2. Screenshots carousel, 3. Features with icons, 4. Reviews/ratings, 5. Download CTAs",Download buttons prominent (App Store + Play Store) throughout,Dark/light matching app store feel. Star ratings in gold. Screenshots with device frames.,"Device mockup rotations, screenshot slider, star rating animations, download button pulse",Show real screenshots. Include ratings (4.5+ stars). QR code for mobile. Platform-specific CTAs.
16,FAQ/Documentation Landing,"faq, documentation, help, support, questions","1. Hero with search bar, 2. Popular categories, 3. FAQ accordion, 4. Contact/support CTA",Search bar prominent + Contact CTA for unresolved questions,"Clean, high readability. Minimal color. Category icons in brand color. Success green for resolved.","Search autocomplete, smooth accordion open/close, category hover, helpful feedback buttons",Reduce support tickets. Track search analytics. Show related articles. Contact escalation path.
17,Immersive/Interactive Experience,"immersive, interactive, experience, 3d, animation","1. Full-screen interactive element, 2. Guided product tour, 3. Key benefits revealed, 4. CTA after completion",After interaction complete + Skip option for impatient users,Immersive experience colors. Dark background for focus. Highlight interactive elements.,"WebGL, 3D interactions, gamification elements, progress indicators, reward animations",40% higher engagement. Performance trade-off. Provide skip option. Mobile fallback essential.
18,Event/Conference Landing,"event, conference, meetup, registration, schedule","1. Hero (date/location/countdown), 2. Speakers grid, 3. Agenda/schedule, 4. Sponsors, 5. Register CTA",Register CTA sticky + After speakers + Bottom,Urgency colors (countdown). Event branding. Speaker cards professional. Sponsor logos neutral.,"Countdown timer, speaker hover cards with bio, agenda tabs, early bird countdown",Early bird pricing with deadline. Social proof (past attendees). Speaker credibility. Multi-ticket discounts.
19,Product Review/Ratings Focused,"reviews, ratings, testimonials, social-proof, stars","1. Hero (product + aggregate rating), 2. Rating breakdown, 3. Individual reviews, 4. Buy/CTA",After reviews summary + Buy button alongside reviews,Trust colors. Star ratings gold. Verified badge green. Review sentiment colors.,"Star fill animations, review filtering, helpful vote interactions, photo lightbox",User-generated content builds trust. Show verified purchases. Filter by rating. Respond to negative reviews.
20,Community/Forum Landing,"community, forum, social, members, discussion","1. Hero (community value prop), 2. Popular topics/categories, 3. Active members showcase, 4. Join CTA",Join button prominent + After member showcase,"Warm, welcoming. Member photos add humanity. Topic badges in brand colors. Activity indicators green.","Member avatars animation, activity feed live updates, topic hover previews, join success celebration","Show active community (member count, posts today). Highlight benefits. Preview content. Easy onboarding."
21,Before-After Transformation,"before-after, transformation, results, comparison","1. Hero (problem state), 2. Transformation slider/comparison, 3. How it works, 4. Results CTA",After transformation reveal + Bottom,Contrast: muted/grey (before) vs vibrant/colorful (after). Success green for results.,"Slider comparison interaction, before/after reveal animations, result counters, testimonial videos",Visual proof of value. 45% higher conversion. Real results. Specific metrics. Guarantee offer.
22,Marketplace / Directory,"marketplace, directory, search, listing","1. Hero (Search focused), 2. Categories, 3. Featured Listings, 4. Trust/Safety, 5. CTA (Become a host/seller)",Hero Search Bar + Navbar 'List your item',Search: High contrast. Categories: Visual icons. Trust: Blue/Green.,Search autocomplete animation," map hover pins, card carousel, Search bar is the CTA. Reduce friction to search. Popular searches suggestions."
23,Newsletter / Content First,"newsletter, content, writer, blog, subscribe","1. Hero (Value Prop + Form), 2. Recent Issues/Archives, 3. Social Proof (Subscriber count), 4. About Author",Hero inline form + Sticky header form,Minimalist. Paper-like background. Text focus. Accent color for Subscribe.,Text highlight animations," typewriter effect, subtle fade-in, Single field form (Email only). Show 'Join X, 000 readers'. Read sample link."
24,Webinar Registration,"webinar, registration, event, training, live","1. Hero (Topic + Timer + Form), 2. What you'll learn, 3. Speaker Bio, 4. Urgency/Bonuses, 5. Form (again)",Hero (Right side form) + Bottom anchor,Urgency: Red/Orange. Professional: Blue/Navy. Form: High contrast white.,Countdown timer," speaker avatar float, urgent ticker, Limited seats logic. 'Live' indicator. Auto-fill timezone."
25,Enterprise Gateway,"enterprise, corporate, gateway, solutions, portal","1. Hero (Video/Mission), 2. Solutions by Industry, 3. Solutions by Role, 4. Client Logos, 5. Contact Sales",Contact Sales (Primary) + Login (Secondary),Corporate: Navy/Grey. High integrity. Conservative accents.,Slow video background," logo carousel, tab switching for industries, Path selection (I am a...). Mega menu navigation. Trust signals prominent."
26,Portfolio Grid,"portfolio, grid, showcase, gallery, masonry","1. Hero (Name/Role), 2. Project Grid (Masonry), 3. About/Philosophy, 4. Contact",Project Card Hover + Footer Contact,Neutral background (let work shine). Text: Black/White. Accent: Minimal.,Image lazy load reveal," hover overlay info, lightbox view, Visuals first. Filter by category. Fast loading essential."
27,Horizontal Scroll Journey,"horizontal, scroll, journey, gallery, storytelling, panoramic","1. Intro (Vertical), 2. The Journey (Horizontal Track), 3. Detail Reveal, 4. Vertical Footer",Floating Sticky CTA or End of Horizontal Track,Continuous palette transition. Chapter colors. Progress bar #000000.,"Scroll-jacking (careful), parallax layers, horizontal slide, progress indicator","Immersive product discovery. High engagement. Keep navigation visible.
28,Bento Grid Showcase,bento, grid, features, modular, apple-style, showcase"", 1. Hero, 2. Bento Grid (Key Features), 3. Detail Cards, 4. Tech Specs, 5. CTA, Floating Action Button or Bottom of Grid, Card backgrounds: #F5F5F7 or Glass. Icons: Vibrant brand colors. Text: Dark., Hover card scale (1.02), video inside cards, tilt effect, staggered reveal, Scannable value props. High information density without clutter. Mobile stack.
29,Interactive 3D Configurator,3d, configurator, customizer, interactive, product"", 1. Hero (Configurator), 2. Feature Highlight (synced), 3. Price/Specs, 4. Purchase, Inside Configurator UI + Sticky Bottom Bar, Neutral studio background. Product: Realistic materials. UI: Minimal overlay., Real-time rendering, material swap animation, camera rotate/zoom, light reflection, Increases ownership feeling. 360 view reduces return rates. Direct add-to-cart.
30,AI-Driven Dynamic Landing,ai, dynamic, personalized, adaptive, generative"", 1. Prompt/Input Hero, 2. Generated Result Preview, 3. How it Works, 4. Value Prop, Input Field (Hero) + 'Try it' Buttons, Adaptive to user input. Dark mode for compute feel. Neon accents., Typing text effects, shimmering generation loaders, morphing layouts, Immediate value demonstration. 'Show, don't tell'. Low friction start."
1 No Pattern Name Keywords Section Order Primary CTA Placement Color Strategy Recommended Effects Conversion Optimization
2 1 Hero + Features + CTA hero, hero-centric, features, feature-rich, cta, call-to-action 1. Hero with headline/image, 2. Value prop, 3. Key features (3-5), 4. CTA section, 5. Footer Hero (sticky) + Bottom Hero: Brand primary or vibrant. Features: Card bg #FAFAFA. CTA: Contrasting accent color Hero parallax, feature card hover lift, CTA glow on hover Deep CTA placement. Use contrasting color (at least 7:1 contrast ratio). Sticky navbar CTA.
3 2 Hero + Testimonials + CTA hero, testimonials, social-proof, trust, reviews, cta 1. Hero, 2. Problem statement, 3. Solution overview, 4. Testimonials carousel, 5. CTA Hero (sticky) + Post-testimonials Hero: Brand color. Testimonials: Light bg #F5F5F5. Quotes: Italic, muted color #666. CTA: Vibrant Testimonial carousel slide animations, quote marks animations, avatar fade-in Social proof before CTA. Use 3-5 testimonials. Include photo + name + role. CTA after social proof.
4 3 Product Demo + Features demo, product-demo, features, showcase, interactive 1. Hero, 2. Product video/mockup (center), 3. Feature breakdown per section, 4. Comparison (optional), 5. CTA Video center + CTA right/bottom Video surround: Brand color overlay. Features: Icon color #0080FF. Text: Dark #222 Video play button pulse, feature scroll reveals, demo interaction highlights Embedded product demo increases engagement. Use interactive mockup if possible. Auto-play video muted.
5 4 Minimal Single Column minimal, simple, direct, single-column, clean 1. Hero headline, 2. Short description, 3. Benefit bullets (3 max), 4. CTA, 5. Footer Center, large CTA button Minimalist: Brand + white #FFFFFF + accent. Buttons: High contrast 7:1+. Text: Black/Dark grey Minimal hover effects. Smooth scroll. CTA scale on hover (subtle) Single CTA focus. Large typography. Lots of whitespace. No nav clutter. Mobile-first.
6 5 Funnel (3-Step Conversion) funnel, conversion, steps, wizard, onboarding 1. Hero, 2. Step 1 (problem), 3. Step 2 (solution), 4. Step 3 (action), 5. CTA progression Each step: mini-CTA. Final: main CTA Step colors: 1 (Red/Problem), 2 (Orange/Process), 3 (Green/Solution). CTA: Brand color Step number animations, progress bar fill, step transitions smooth scroll Progressive disclosure. Show only essential info per step. Use progress indicators. Multiple CTAs.
7 6 Comparison Table + CTA comparison, table, compare, versus, cta 1. Hero, 2. Problem intro, 3. Comparison table (product vs competitors), 4. Pricing (optional), 5. CTA Table: Right column. CTA: Below table Table: Alternating rows (white/light grey). Your product: Highlight #FFFACD (light yellow) or green. Text: Dark Table row hover highlight, price toggle animations, feature checkmark animations Use comparison to show unique value. Highlight your product row. Include 'free trial' in pricing row.
8 7 Lead Magnet + Form lead, form, signup, capture, email, magnet 1. Hero (benefit headline), 2. Lead magnet preview (ebook cover, checklist, etc), 3. Form (minimal fields), 4. CTA submit Form CTA: Submit button Lead magnet: Professional design. Form: Clean white bg. Inputs: Light border #CCCCCC. CTA: Brand color Form focus state animations, input validation animations, success confirmation animation Form fields ≤ 3 for best conversion. Offer valuable lead magnet preview. Show form submission progress.
9 8 Pricing Page + CTA pricing, plans, tiers, comparison, cta 1. Hero (pricing headline), 2. Price comparison cards, 3. Feature comparison table, 4. FAQ section, 5. Final CTA Each card: CTA button. Sticky CTA in nav Free: Grey, Starter: Blue, Pro: Green/Gold, Enterprise: Dark. Cards: 1px border, shadow Price toggle animation (monthly/yearly), card comparison highlight, FAQ accordion open/close Recommend starter plan (pre-select/highlight). Show annual discount (20-30%). Use FAQs to address concerns.
10 9 Video-First Hero video, hero, media, visual, engaging 1. Hero with video background, 2. Key features overlay, 3. Benefits section, 4. CTA Overlay on video (center/bottom) + Bottom section Dark overlay 60% on video. Brand accent for CTA. White text on dark. Video autoplay muted, parallax scroll, text fade-in on scroll 86% higher engagement with video. Add captions for accessibility. Compress video for performance.
11 10 Scroll-Triggered Storytelling storytelling, scroll, narrative, story, immersive 1. Intro hook, 2. Chapter 1 (problem), 3. Chapter 2 (journey), 4. Chapter 3 (solution), 5. Climax CTA End of each chapter (mini) + Final climax CTA Progressive reveal. Each chapter has distinct color. Building intensity. ScrollTrigger animations, parallax layers, progressive disclosure, chapter transitions Narrative increases time-on-page 3x. Use progress indicator. Mobile: simplify animations.
12 11 AI Personalization Landing ai, personalization, smart, recommendation, dynamic 1. Dynamic hero (personalized), 2. Relevant features, 3. Tailored testimonials, 4. Smart CTA Context-aware placement based on user segment Adaptive based on user data. A/B test color variations per segment. Dynamic content swap, fade transitions, personalized product recommendations 20%+ conversion with personalization. Requires analytics integration. Fallback for new users.
13 12 Waitlist/Coming Soon waitlist, coming-soon, launch, early-access, notify 1. Hero with countdown, 2. Product teaser/preview, 3. Email capture form, 4. Social proof (waitlist count) Email form prominent (above fold) + Sticky form on scroll Anticipation: Dark + accent highlights. Countdown in brand color. Urgency indicators. Countdown timer animation, email validation feedback, success confetti, social share buttons Scarcity + exclusivity. Show waitlist count. Early access benefits. Referral program.
14 13 Comparison Table Focus comparison, table, versus, compare, features 1. Hero (problem statement), 2. Comparison matrix (you vs competitors), 3. Feature deep-dive, 4. Winner CTA After comparison table (highlighted row) + Bottom Your product column highlighted (accent bg or green). Competitors neutral. Checkmarks green. Table row hover highlight, feature checkmark animations, sticky comparison header Show value vs competitors. 35% higher conversion. Be factual. Include pricing if favorable.
15 14 Pricing-Focused Landing pricing, price, cost, plans, subscription 1. Hero (value proposition), 2. Pricing cards (3 tiers), 3. Feature comparison, 4. FAQ, 5. Final CTA Each pricing card + Sticky CTA in nav + Bottom Popular plan highlighted (brand color border/bg). Free: grey. Enterprise: dark/premium. Price toggle monthly/annual animation, card hover lift, FAQ accordion smooth open Annual discount 20-30%. Recommend mid-tier (most popular badge). Address objections in FAQ.
16 15 App Store Style Landing app, mobile, download, store, install 1. Hero with device mockup, 2. Screenshots carousel, 3. Features with icons, 4. Reviews/ratings, 5. Download CTAs Download buttons prominent (App Store + Play Store) throughout Dark/light matching app store feel. Star ratings in gold. Screenshots with device frames. Device mockup rotations, screenshot slider, star rating animations, download button pulse Show real screenshots. Include ratings (4.5+ stars). QR code for mobile. Platform-specific CTAs.
17 16 FAQ/Documentation Landing faq, documentation, help, support, questions 1. Hero with search bar, 2. Popular categories, 3. FAQ accordion, 4. Contact/support CTA Search bar prominent + Contact CTA for unresolved questions Clean, high readability. Minimal color. Category icons in brand color. Success green for resolved. Search autocomplete, smooth accordion open/close, category hover, helpful feedback buttons Reduce support tickets. Track search analytics. Show related articles. Contact escalation path.
18 17 Immersive/Interactive Experience immersive, interactive, experience, 3d, animation 1. Full-screen interactive element, 2. Guided product tour, 3. Key benefits revealed, 4. CTA after completion After interaction complete + Skip option for impatient users Immersive experience colors. Dark background for focus. Highlight interactive elements. WebGL, 3D interactions, gamification elements, progress indicators, reward animations 40% higher engagement. Performance trade-off. Provide skip option. Mobile fallback essential.
19 18 Event/Conference Landing event, conference, meetup, registration, schedule 1. Hero (date/location/countdown), 2. Speakers grid, 3. Agenda/schedule, 4. Sponsors, 5. Register CTA Register CTA sticky + After speakers + Bottom Urgency colors (countdown). Event branding. Speaker cards professional. Sponsor logos neutral. Countdown timer, speaker hover cards with bio, agenda tabs, early bird countdown Early bird pricing with deadline. Social proof (past attendees). Speaker credibility. Multi-ticket discounts.
20 19 Product Review/Ratings Focused reviews, ratings, testimonials, social-proof, stars 1. Hero (product + aggregate rating), 2. Rating breakdown, 3. Individual reviews, 4. Buy/CTA After reviews summary + Buy button alongside reviews Trust colors. Star ratings gold. Verified badge green. Review sentiment colors. Star fill animations, review filtering, helpful vote interactions, photo lightbox User-generated content builds trust. Show verified purchases. Filter by rating. Respond to negative reviews.
21 20 Community/Forum Landing community, forum, social, members, discussion 1. Hero (community value prop), 2. Popular topics/categories, 3. Active members showcase, 4. Join CTA Join button prominent + After member showcase Warm, welcoming. Member photos add humanity. Topic badges in brand colors. Activity indicators green. Member avatars animation, activity feed live updates, topic hover previews, join success celebration Show active community (member count, posts today). Highlight benefits. Preview content. Easy onboarding.
22 21 Before-After Transformation before-after, transformation, results, comparison 1. Hero (problem state), 2. Transformation slider/comparison, 3. How it works, 4. Results CTA After transformation reveal + Bottom Contrast: muted/grey (before) vs vibrant/colorful (after). Success green for results. Slider comparison interaction, before/after reveal animations, result counters, testimonial videos Visual proof of value. 45% higher conversion. Real results. Specific metrics. Guarantee offer.
23 22 Marketplace / Directory marketplace, directory, search, listing 1. Hero (Search focused), 2. Categories, 3. Featured Listings, 4. Trust/Safety, 5. CTA (Become a host/seller) Hero Search Bar + Navbar 'List your item' Search: High contrast. Categories: Visual icons. Trust: Blue/Green. Search autocomplete animation map hover pins, card carousel, Search bar is the CTA. Reduce friction to search. Popular searches suggestions.
24 23 Newsletter / Content First newsletter, content, writer, blog, subscribe 1. Hero (Value Prop + Form), 2. Recent Issues/Archives, 3. Social Proof (Subscriber count), 4. About Author Hero inline form + Sticky header form Minimalist. Paper-like background. Text focus. Accent color for Subscribe. Text highlight animations typewriter effect, subtle fade-in, Single field form (Email only). Show 'Join X, 000 readers'. Read sample link.
25 24 Webinar Registration webinar, registration, event, training, live 1. Hero (Topic + Timer + Form), 2. What you'll learn, 3. Speaker Bio, 4. Urgency/Bonuses, 5. Form (again) Hero (Right side form) + Bottom anchor Urgency: Red/Orange. Professional: Blue/Navy. Form: High contrast white. Countdown timer speaker avatar float, urgent ticker, Limited seats logic. 'Live' indicator. Auto-fill timezone.
26 25 Enterprise Gateway enterprise, corporate, gateway, solutions, portal 1. Hero (Video/Mission), 2. Solutions by Industry, 3. Solutions by Role, 4. Client Logos, 5. Contact Sales Contact Sales (Primary) + Login (Secondary) Corporate: Navy/Grey. High integrity. Conservative accents. Slow video background logo carousel, tab switching for industries, Path selection (I am a...). Mega menu navigation. Trust signals prominent.
27 26 Portfolio Grid portfolio, grid, showcase, gallery, masonry 1. Hero (Name/Role), 2. Project Grid (Masonry), 3. About/Philosophy, 4. Contact Project Card Hover + Footer Contact Neutral background (let work shine). Text: Black/White. Accent: Minimal. Image lazy load reveal hover overlay info, lightbox view, Visuals first. Filter by category. Fast loading essential.
28 27 Horizontal Scroll Journey horizontal, scroll, journey, gallery, storytelling, panoramic 1. Intro (Vertical), 2. The Journey (Horizontal Track), 3. Detail Reveal, 4. Vertical Footer Floating Sticky CTA or End of Horizontal Track Continuous palette transition. Chapter colors. Progress bar #000000. Scroll-jacking (careful), parallax layers, horizontal slide, progress indicator Immersive product discovery. High engagement. Keep navigation visible. 28,Bento Grid Showcase,bento, grid, features, modular, apple-style, showcase", 1. Hero, 2. Bento Grid (Key Features), 3. Detail Cards, 4. Tech Specs, 5. CTA, Floating Action Button or Bottom of Grid, Card backgrounds: #F5F5F7 or Glass. Icons: Vibrant brand colors. Text: Dark., Hover card scale (1.02), video inside cards, tilt effect, staggered reveal, Scannable value props. High information density without clutter. Mobile stack. 29,Interactive 3D Configurator,3d, configurator, customizer, interactive, product", 1. Hero (Configurator), 2. Feature Highlight (synced), 3. Price/Specs, 4. Purchase, Inside Configurator UI + Sticky Bottom Bar, Neutral studio background. Product: Realistic materials. UI: Minimal overlay., Real-time rendering, material swap animation, camera rotate/zoom, light reflection, Increases ownership feeling. 360 view reduces return rates. Direct add-to-cart. 30,AI-Driven Dynamic Landing,ai, dynamic, personalized, adaptive, generative", 1. Prompt/Input Hero, 2. Generated Result Preview, 3. How it Works, 4. Value Prop, Input Field (Hero) + 'Try it' Buttons, Adaptive to user input. Dark mode for compute feel. Neon accents., Typing text effects, shimmering generation loaders, morphing layouts, Immediate value demonstration. 'Show, don't tell'. Low friction start.

Some files were not shown because too many files have changed in this diff Show More