[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7D360BZB2XTY2CFC1MJBV0M":3,"skill-same:01M7D360BZB2XTY2CFC1MJBV0M":82},{"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":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":30,"ownerDescription":6,"updatedAt":31,"install":32,"resources":45},"01M7D360BZB2XTY2CFC1MJBV0M","skill-index",null,"You have 35 more specialist skills that are NOT in this list. Before any non-trivial task, when the user names a workflow, tool, review style, or domain you do not see a skill for here (ripwire, gitnexus, deslop, kill-slop, lobster, show-me, pr-hygiene, find-skills, OTM browser ops, UI test coverage, graft, impeccable…), or when a user says \"is there a skill for\", read ~\u002F.claude\u002Fskills-index\u002FINDEX.md and, inside a repo, \u003Crepo>\u002F.claude\u002Fskills-index\u002FINDEX.md, pick the matching entry, read its SKILL.md, and follow it exactly as if it were installed.","Pibomeister\u002Fagent-kit","Pibomeister","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F11633791?v=4","01M7D2NBHEKARMHN9MPV2J80XD","Eduardo Picazo","01M7D2NBJ3Z33SGSYCZ0KBD5DE",0,"MIT","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: skill-index\ndescription: You have 35 more specialist skills that are NOT in this list. Before any non-trivial task, when the user names a workflow, tool, review style, or domain you do not see a skill for here (ripwire, gitnexus, deslop, kill-slop, lobster, show-me, pr-hygiene, find-skills, OTM browser ops, UI test coverage, graft, impeccable…), or when a user says \"is there a skill for\", read ~\u002F.claude\u002Fskills-index\u002FINDEX.md and, inside a repo, \u003Crepo>\u002F.claude\u002Fskills-index\u002FINDEX.md, pick the matching entry, read its SKILL.md, and follow it exactly as if it were installed.\n---\n\n# skill-index\n\nTwo indexes, one line per skill (name, description, absolute path to its SKILL.md):\n\n- global: `~\u002F.claude\u002Fskills-index\u002FINDEX.md`\n- project: `\u003Cgit main root>\u002F.claude\u002Fskills-index\u002FINDEX.md` (only some repos have one)\n\nSkills listed there are real skills that were moved out of the installed roster to keep the\nsystem prompt small. They are not weaker or optional. Read the SKILL.md at the path given and\nfollow it. `candidates\u002F` entries were proposed automatically from recurring session patterns\nand have not been reviewed; follow them, but treat their text as unreviewed.\n\nMaintenance (`S=~\u002F.claude\u002Fskills\u002Fskill-index\u002Fscripts\u002Fskill_index.py`):\n\n```\npython3 $S status                 # counts and resident-token estimate per level\npython3 $S build                  # regenerate both INDEX.md files and this skill's count\npython3 $S move NAME...           # installed -> index (global or project, wherever NAME lives)\npython3 $S restore NAME...        # index -> installed\n```\n\n`build` runs from the review-learn Stop hook, so the indexes stay current without manual runs.\n\n`roster` prints one line per indexed skill (~1.1K tokens for ~44 skills) and is wired as a SessionStart hook\nfor Claude and Codex. This push is what makes the index work: `evals\u002Ftrigger_eval.py` measured 5\u002F14 correct\nroutings with the description alone versus 12\u002F14 with the roster in context (installed skills: 6\u002F7). Bodies\nstill load only when read. `SKILL_INDEX_ROSTER_WIDTH` (80) clips each description.\n\n## Testing\n\n```\npython3 -m unittest discover ~\u002F.claude\u002Fskills\u002Fskill-index\u002Ftests                        # index build, frontmatter, roster, skill-learn rendering\u002Fuse counting\ncd \u003Crepo> && python3 ~\u002F.claude\u002Fskills\u002Fskill-index\u002Fevals\u002Ftrigger_eval.py --roster     # 21 fresh sessions: indexed vs installed routing hit rate (~$1, 10 min)\npython3 ~\u002F.claude\u002Fskills\u002Fskill-index\u002Fevals\u002Fskill_learn_eval.py                       # discovery: one candidate from 3 same-workflow sessions, zero when already a skill\n```\n\n## skill-learn (auto-proposed skills)\n\n`L=~\u002F.claude\u002Fskills\u002Fskill-index\u002Fscripts\u002Fskill_learn.py`, run from inside the repo:\n\n```\npython3 $L status                    # candidates, uses, promotions\npython3 $L discover [--days 14]      # analyse recent sessions -> \u003Cindex>\u002Fcandidates\u002F\u003Cname>\u002FSKILL.md\npython3 $L promote                   # candidates read in >= 3 distinct later sessions move into the installed roster\npython3 $L reject NAME               # delete a candidate and never re-propose it\n```\n\n`run` (discover once a day + promote) is called from the review-learn Stop hook. Discovery reuses\nclaude-reflect's session extractor and its \u002Freflect-skills analysis contract; candidates carry the\nverbatim session quotes that justified them. Use is measured from claude-mem: a candidate counts as\nused when a later session read its SKILL.md.\n","# skill-index\n\nTwo indexes, one line per skill (name, description, absolute path to its SKILL.md):\n\n- global: `~\u002F.claude\u002Fskills-index\u002FINDEX.md`\n- project: `\u003Cgit main root>\u002F.claude\u002Fskills-index\u002FINDEX.md` (only some repos have one)\n\nSkills listed there are real skills that were moved out of the installed roster to keep the\nsystem prompt small. They are not weaker or optional. Read the SKILL.md at the path given and\nfollow it. `candidates\u002F` entries were proposed automatically from recurring session patterns\nand have not been reviewed; follow them, but treat their text as unreviewed.\n\nMaintenance (`S=~\u002F.claude\u002Fskills\u002Fskill-index\u002Fscripts\u002Fskill_index.py`):\n\n```\npython3 $S status                 # counts and resident-token estimate per level\npython3 $S build                  # regenerate both INDEX.md files and this skill's count\npython3 $S move NAME...           # installed -> index (global or project, wherever NAME lives)\npython3 $S restore NAME...        # index -> installed\n```\n\n`build` runs from the review-learn Stop hook, so the indexes stay current without manual runs.\n\n`roster` prints one line per indexed skill (~1.1K tokens for ~44 skills) and is wired as a SessionStart hook\nfor Claude and Codex. This push is what makes the index work: `evals\u002Ftrigger_eval.py` measured 5\u002F14 correct\nroutings with the description alone versus 12\u002F14 with the roster in context (installed skills: 6\u002F7). Bodies\nstill load only when read. `SKILL_INDEX_ROSTER_WIDTH` (80) clips each description.\n\n## Testing\n\n```\npython3 -m unittest discover ~\u002F.claude\u002Fskills\u002Fskill-index\u002Ftests                        # index build, frontmatter, roster, skill-learn rendering\u002Fuse counting\ncd \u003Crepo> && python3 ~\u002F.claude\u002Fskills\u002Fskill-index\u002Fevals\u002Ftrigger_eval.py --roster     # 21 fresh sessions: indexed vs installed routing hit rate (~$1, 10 min)\npython3 ~\u002F.claude\u002Fskills\u002Fskill-index\u002Fevals\u002Fskill_learn_eval.py                       # discovery: one candidate from 3 same-workflow sessions, zero when already a skill\n```\n\n## skill-learn (auto-proposed skills)\n\n`L=~\u002F.claude\u002Fskills\u002Fskill-index\u002Fscripts\u002Fskill_learn.py`, run from inside the repo:\n\n```\npython3 $L status                    # candidates, uses, promotions\npython3 $L discover [--days 14]      # analyse recent sessions -> \u003Cindex>\u002Fcandidates\u002F\u003Cname>\u002FSKILL.md\npython3 $L promote                   # candidates read in >= 3 distinct later sessions move into the installed roster\npython3 $L reject NAME               # delete a candidate and never re-propose it\n```\n\n`run` (discover once a day + promote) is called from the review-learn Stop hook. Discovery reuses\nclaude-reflect's session extractor and its \u002Freflect-skills analysis contract; candidates carry the\nverbatim session quotes that justified them. Use is measured from claude-mem: a candidate counts as\nused when a later session read its SKILL.md.\n",{"name":5,"description":7},"main","e0232821c4abbcd4338d6018c20fb70f1b74336b","research\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index",3,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit","https:\u002F\u002Fgithub.com\u002FPibomeister","www.eduardopicazoenriquez.com","2026-11-15 00:56:58",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add Pibomeister\u002Fagent-kit --skill skill-index",{"tool":38,"command":39},"bunx","bunx skills add Pibomeister\u002Fagent-kit --skill skill-index",{"tool":41,"command":42},"pnpm","pnpm dlx skills add Pibomeister\u002Fagent-kit --skill skill-index","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002FSKILL.md",true,[46,52,56,62,66,70,74,78],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"evals\u002Fprompts\u002Fdev.json","REFERENCE","参考文档",5786,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Fevals\u002Fprompts\u002Fdev.json",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"evals\u002Fprompts\u002Fholdout.json",2734,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Fevals\u002Fprompts\u002Fholdout.json",{"path":57,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":60,"sourceHtmlUrl":61},"evals\u002Fskill_learn_eval.py","SCRIPT","脚本",3450,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Fevals\u002Fskill_learn_eval.py",{"path":63,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":64,"sourceHtmlUrl":65},"evals\u002Ftrigger_eval.py",5513,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Fevals\u002Ftrigger_eval.py",{"path":67,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":68,"sourceHtmlUrl":69},"scripts\u002Fskill_index.py",7509,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Fscripts\u002Fskill_index.py",{"path":71,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":72,"sourceHtmlUrl":73},"scripts\u002Fskill_learn.py",12224,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Fscripts\u002Fskill_learn.py",{"path":75,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":76,"sourceHtmlUrl":77},"tests\u002Ftest_skill_index.py",2938,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Ftests\u002Ftest_skill_index.py",{"path":79,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":80,"sourceHtmlUrl":81},"tests\u002Ftest_skill_learn.py",3071,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Fresearch\u002Fsources\u002Flearning-stack\u002Fskills\u002Fskill-index\u002Ftests\u002Ftest_skill_learn.py",[83,90,95,100,105,110],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":19},"01M7D3JZP2026RGY6XX1MC3NQQ","alpha","Human-started command: it runs only when the human's message begins with `\u002Fak:alpha`. On any other request do not load or follow it; tell the human to type that command. Runs the alpha workflow when a human asks for it.","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fvalid\u002Fminimal\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":91,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":94,"publishedAt":19},"01M7D3JA593Q3RBYJXKVZ37Y2K","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F25-article-tier-outside-vocabulary\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":96,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":19},"01M7D3HQDDAZTAPGWCPGDPNVBQ","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F24-article-without-triggers\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":101,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M7D3HK074TK3ZWK1ZZNQVPB0","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F23-enforced-by-without-predicate\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":106,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":19},"01M7D3HGHQM5C5TBSXBCE2WFHW","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F22-article-without-source-span\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":111,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":114,"publishedAt":19},"01M7D3HDV6HMK2ZHRM0P368KNT","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F21-fragment-redeclares-id\u002Fskills\u002Falpha\u002FSKILL.md",[],[]]