[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72T8902P29X843B3K9YR4FN":3,"skill-same:01M72T8902P29X843B3K9YR4FN":45},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":27,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":29,"updatedAt":30,"install":31,"resources":44},"01M72T8902P29X843B3K9YR4FN","worldexams-quality-gate",null,"Use when validating or repairing generated WorldExams\u002FSaberParaTodos bundle markdown before opening a PR. CJK\u002FEnglish leakage, evidenced ICFES axes, answer-letter rebalance, dual validators.","iberi22\u002Fskills","iberi22","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F10615454?v=4","01M72SQ2NXAA3GW0FSH053A5ES","Brahyan Belalcazar","01M72SQ2PYCH0SSA1K61NGEHDV",0,"MIT","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fgeneral\u002Fworldexams-quality-gate\u002FSKILL.md",[],[],"2026-11-10 22:39:42",false,"---\nname: worldexams-quality-gate\ndescription: Use when validating or repairing generated WorldExams\u002FSaberParaTodos bundle markdown before opening a PR. CJK\u002FEnglish leakage, evidenced ICFES axes, answer-letter rebalance, dual validators.\n---\n\n# WorldExams quality gate v2\n\nValidates generated `questions_data\u002F**\u002Fweekly\u002F*-MASTERY-bundle.md` before it reaches a PR.\nCalibrated against already-merged content: a gate that flags merged files is mis-specified, not the content.\n\n## Two validators, both must pass\n\n```bash\nnode scripts\u002Fvalidate-bundles-v52.mjs \u003Cfile.md> [more.md ...]\nnode saberparatodos\u002Fscripts\u002Fvalidate_content.js --country=colombia --grade=\u003CN>  # needs node_modules\n```\n\nWHERE EACH ONE ACTUALLY RUNS, verified 2026-10-01:\n\n- `validate-bundles-v52.mjs` runs in **CI** (`.github\u002Fworkflows\u002Fdeploy-preview.yml`),\n  on the changed bundle files. It is `npm run validate`.\n- It does **NOT** run in `.husky\u002Fpre-commit`. The hook runs three things: the\n  SWAL leak gate, `scripts\u002Fhusky-guard.mjs --staged`, and the interlock that\n  blocks unsafe generators. A bundle with bad feedback commits locally and is\n  caught by CI, not by the hook.\n- `validate_content.js` is a separate older validator under `saberparatodos\u002F`.\n  It is not wired into the hook or into `deploy-preview.yml`.\n\nSo a green `git commit` does not mean the content passed. Run the validator\nyourself before pushing, or wait for CI.\n\n`validate_content.js` needs `saberparatodos\u002Fnode_modules`. In a fresh worktree, symlink it from the\nmain checkout, run, then delete the symlink. To validate files that only exist in a worktree, copy\nthem into the main checkout, run, then **delete the copies**.\n\n## Defects the generator actually produces\n\n| Defect | Detection | Fix |\n|---|---|---|\n| CJK injected into Spanish prose | `grep -nP '[\\x{4e00}-\\x{9fff}\\x{3040}-\\x{30ff}\\x{ac00}-\\x{d7af}]'` | rewrite the phrase in Spanish |\n| Cyrillic\u002FGreek in words | `grep -nP '[\\x{0400}-\\x{04ff}\\x{0370}-\\x{03ff}]'` | transliterate or rewrite |\n| English words in Spanish prose (`signalling`, `el_noise`) | word list over slug-stripped text | replace with Spanish |\n| Duplicated `**Contexto:**` line | count each field vs question count | drop the 2nd (mangled) copy |\n| A question missing `**ICFES:**` | `ICFES count != question count` | restore the evidenced axis |\n| Answer-letter bias (7\u002F12 on B, or zero D) | repo validator warning | rebalance (below) |\n\n**The CJK grep range in the cron spec is malformed**: `\\x{ac00}-\\xd7af` errors with\n\"range out of order in character class\". The working endpoint is `\\x{d7af}`.\nAlways confirm a detector detects: append real CJK to a temp copy and grep it.\n\n## Accent check must skip slugs\n\nProtocol requires ASCII kebab slugs and `**ID:**` values, so `el-realismo-magico` is CORRECT and\n`Realismo Mágico` in prose is also correct. Strip slugs first, then look for the unaccented word:\n\n```python\nre.sub(r\"[a-z0-9]+(?:-[a-z0-9]+)+\", \" \", line)   # remove slugs\u002Fids\nre.search(r\"\\b(magico|tambien|fisica|quimica|cancer|farmacos)\\b\", stripped)\n```\n\nDo NOT add real Spanish cognates to the English blocklist (`defender`, `correct`, `answer` are\nvalid Spanish and appear in merged content).\n\n## ICFES axes must be evidenced\n\nAn axis is valid only if it already appears in >=3 merged bundles of the SAME subject. Compute the\nset from the subject dir rather than guessing. For `lengua` the evidenced trio is:\n\n- `Competencia Lectora (Literal)`\n- `Competencia Lectora (Inferencial)`\n- `Competencia Lectora (Crítica)`\n\nForbidden: the subject name as the axis, combined forms, and axes from another subject.\n\n## Difficulty ranges\n\n`validate_content.js` only requires each header be in the ALLOWED SET\n`{D3-D4, D5-D6, D7-D8, D9-D10}` - it does NOT enforce a fixed distribution. A 3\u002F3\u002F4\u002F2 split is\ntolerated in merged content. Enforce monotonic (non-decreasing) progression as an error; treat\ndistribution drift as a warning only.\n\n## Answer-letter rebalance\n\nRotate the correct option as a WHOLE LINE so its `\u003C!-- feedback: -->` travels with it, then\nrelabel A)-D). Never rewrite option text. Target 3\u002F3\u002F3\u002F3 in a 12-question bundle; move only from\nover-represented letters to under-represented ones. Verify on a copy first: option text + feedback\nsets must be identical per question before\u002Fafter.\n\n## Static packs\n\nNever hand-edit packs. Regenerate from the markdown:\n\n```bash\nnode saberparatodos\u002Fscripts\u002Fgenerate-static-packs.js --all-weekly --changed-only\n```\n\n`.husky\u002Fpre-push` calls `scripts\u002Fhusky-guard.mjs --changed`, which regenerates the\nchanged packs itself and then calls `assertGeneratedPacksCommitted()`. So a\ncommit that leaves a pack unstaged is refused at push time, not at commit time.\nAfter generating by hand, verify each pack's `correct_answer` matches the\nmarkdown `[x]` letter and that every `options[].feedback` is populated.\n\n## Commit hygiene\n\nThe pre-commit hook validates only STAGED bundles, so stage the verified subset explicitly rather\nthan `git add questions_data\u002F`. A failing W31 in the tree must not block a good W27-W30 commit.\n\n## Related\n\n- [[hermes-cron-worldexams-pipeline]] - the 30-min autonomous tick\n","# WorldExams quality gate v2\n\nValidates generated `questions_data\u002F**\u002Fweekly\u002F*-MASTERY-bundle.md` before it reaches a PR.\nCalibrated against already-merged content: a gate that flags merged files is mis-specified, not the content.\n\n## Two validators, both must pass\n\n```bash\nnode scripts\u002Fvalidate-bundles-v52.mjs \u003Cfile.md> [more.md ...]\nnode saberparatodos\u002Fscripts\u002Fvalidate_content.js --country=colombia --grade=\u003CN>  # needs node_modules\n```\n\nWHERE EACH ONE ACTUALLY RUNS, verified 2026-10-01:\n\n- `validate-bundles-v52.mjs` runs in **CI** (`.github\u002Fworkflows\u002Fdeploy-preview.yml`),\n  on the changed bundle files. It is `npm run validate`.\n- It does **NOT** run in `.husky\u002Fpre-commit`. The hook runs three things: the\n  SWAL leak gate, `scripts\u002Fhusky-guard.mjs --staged`, and the interlock that\n  blocks unsafe generators. A bundle with bad feedback commits locally and is\n  caught by CI, not by the hook.\n- `validate_content.js` is a separate older validator under `saberparatodos\u002F`.\n  It is not wired into the hook or into `deploy-preview.yml`.\n\nSo a green `git commit` does not mean the content passed. Run the validator\nyourself before pushing, or wait for CI.\n\n`validate_content.js` needs `saberparatodos\u002Fnode_modules`. In a fresh worktree, symlink it from the\nmain checkout, run, then delete the symlink. To validate files that only exist in a worktree, copy\nthem into the main checkout, run, then **delete the copies**.\n\n## Defects the generator actually produces\n\n| Defect | Detection | Fix |\n|---|---|---|\n| CJK injected into Spanish prose | `grep -nP '[\\x{4e00}-\\x{9fff}\\x{3040}-\\x{30ff}\\x{ac00}-\\x{d7af}]'` | rewrite the phrase in Spanish |\n| Cyrillic\u002FGreek in words | `grep -nP '[\\x{0400}-\\x{04ff}\\x{0370}-\\x{03ff}]'` | transliterate or rewrite |\n| English words in Spanish prose (`signalling`, `el_noise`) | word list over slug-stripped text | replace with Spanish |\n| Duplicated `**Contexto:**` line | count each field vs question count | drop the 2nd (mangled) copy |\n| A question missing `**ICFES:**` | `ICFES count != question count` | restore the evidenced axis |\n| Answer-letter bias (7\u002F12 on B, or zero D) | repo validator warning | rebalance (below) |\n\n**The CJK grep range in the cron spec is malformed**: `\\x{ac00}-\\xd7af` errors with\n\"range out of order in character class\". The working endpoint is `\\x{d7af}`.\nAlways confirm a detector detects: append real CJK to a temp copy and grep it.\n\n## Accent check must skip slugs\n\nProtocol requires ASCII kebab slugs and `**ID:**` values, so `el-realismo-magico` is CORRECT and\n`Realismo Mágico` in prose is also correct. Strip slugs first, then look for the unaccented word:\n\n```python\nre.sub(r\"[a-z0-9]+(?:-[a-z0-9]+)+\", \" \", line)   # remove slugs\u002Fids\nre.search(r\"\\b(magico|tambien|fisica|quimica|cancer|farmacos)\\b\", stripped)\n```\n\nDo NOT add real Spanish cognates to the English blocklist (`defender`, `correct`, `answer` are\nvalid Spanish and appear in merged content).\n\n## ICFES axes must be evidenced\n\nAn axis is valid only if it already appears in >=3 merged bundles of the SAME subject. Compute the\nset from the subject dir rather than guessing. For `lengua` the evidenced trio is:\n\n- `Competencia Lectora (Literal)`\n- `Competencia Lectora (Inferencial)`\n- `Competencia Lectora (Crítica)`\n\nForbidden: the subject name as the axis, combined forms, and axes from another subject.\n\n## Difficulty ranges\n\n`validate_content.js` only requires each header be in the ALLOWED SET\n`{D3-D4, D5-D6, D7-D8, D9-D10}` - it does NOT enforce a fixed distribution. A 3\u002F3\u002F4\u002F2 split is\ntolerated in merged content. Enforce monotonic (non-decreasing) progression as an error; treat\ndistribution drift as a warning only.\n\n## Answer-letter rebalance\n\nRotate the correct option as a WHOLE LINE so its `\u003C!-- feedback: -->` travels with it, then\nrelabel A)-D). Never rewrite option text. Target 3\u002F3\u002F3\u002F3 in a 12-question bundle; move only from\nover-represented letters to under-represented ones. Verify on a copy first: option text + feedback\nsets must be identical per question before\u002Fafter.\n\n## Static packs\n\nNever hand-edit packs. Regenerate from the markdown:\n\n```bash\nnode saberparatodos\u002Fscripts\u002Fgenerate-static-packs.js --all-weekly --changed-only\n```\n\n`.husky\u002Fpre-push` calls `scripts\u002Fhusky-guard.mjs --changed`, which regenerates the\nchanged packs itself and then calls `assertGeneratedPacksCommitted()`. So a\ncommit that leaves a pack unstaged is refused at push time, not at commit time.\nAfter generating by hand, verify each pack's `correct_answer` matches the\nmarkdown `[x]` letter and that every `options[].feedback` is populated.\n\n## Commit hygiene\n\nThe pre-commit hook validates only STAGED bundles, so stage the verified subset explicitly rather\nthan `git add questions_data\u002F`. A failing W31 in the tree must not block a good W27-W30 commit.\n\n## Related\n\n- [[hermes-cron-worldexams-pipeline]] - the 30-min autonomous tick\n",{"name":5,"description":7},"main","688810f2886d212632acf81e532b24e483dd8b98","skills\u002Fgeneral\u002Fworldexams-quality-gate","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills","https:\u002F\u002Fgithub.com\u002Fiberi22","Senior WEB Full-Stack","2026-10-12 02:46:21",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add iberi22\u002Fskills --skill worldexams-quality-gate",{"tool":37,"command":38},"bunx","bunx skills add iberi22\u002Fskills --skill worldexams-quality-gate",{"tool":40,"command":41},"pnpm","pnpm dlx skills add iberi22\u002Fskills --skill worldexams-quality-gate","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fgeneral\u002Fworldexams-quality-gate\u002FSKILL.md",true,[],[46,52,59,65,71,77],{"skillNo":47,"name":48,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":49,"categories":50,"keywords":51,"publishedAt":19},"01M72TWQEQD46ZC9TRBAABF78P","web-design-guidelines","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fweb\u002Fweb-design-guidelines\u002FSKILL.md",[],[],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":19},"01M72TWMQFTYD08DGXVAXRPPEB","blocked-page-recovery","Use when a fetch fails: 403\u002F429, paywall, WAF, bot wall. Recovers the page via archives and provenance-checked fallbacks.","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fweb\u002Fblocked-page-recovery\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":19},"01M72TVQMS3HCTX18C8T97WTQQ","studio-coach","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Fstudio-coach\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M72TVNA2PR9RZW0ZAXYP1G6J","src-generator","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Fsrc-generator\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M72TVEJHHB6SAT3R3TY9MJKS","sqlite-pro","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Fsqlite-pro\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M72TVC2EVBHQHCX96VTTYKB8","rigor-evaluator","Use when you need a quick heuristic rigor score (10 criteria, 0-2 each) for the features in a .gitcore\u002Ffeatures.json file, to decide which features are ready to ship.","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Frigor-evaluator\u002FSKILL.md",[],[]]