[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AAK8C4HZASRCYNFYH4CJF6":3,"skill-same:01M7AAK8C4HZASRCYNFYH4CJF6":59},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":13,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":31,"install":32,"resources":44},"01M7AAK8C4HZASRCYNFYH4CJF6","method-gap-research-status",null,"Read-only family census that recommends one skill, or none, for a target-specific method-gap research census.","joeloverbeck\u002Fskill-evidence","joeloverbeck","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F120873?v=4","01M758KCBJ0GWAPGN74C9NJF8T","01M7AAJ49JAQNVPTMQSWVB57TY",0,"MIT","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmethod-gap-research-status\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: method-gap-research-status\ndescription: Read-only family census that recommends one skill, or none, for a target-specific method-gap research census.\nargument-hint: \"\u003Cskill-name-family, for example game-*>\"\ndisable-model-invocation: true\n---\n\n# Method-Gap Research Status\n\nRun one read-only census across an explicitly named family of skills. Recommend\nat most one exact target for a later, census-only method-gap commission.\nThis skill is a portfolio selector, not a method-gap auditor: it never concludes\nthat a method gap exists, decides that research is due, authors a brief, or\ninvokes the commission.\n\nArguments: one skill-name family expressed as a literal name prefix followed by\none terminal `*`, for example `game-*`. Paths, multiple wildcards, character\nclasses, brace expansion, and a bare `*` are refused.\n\n## Hard boundaries\n\n- Read only: never modify a target, evidence stream, stored projection, research\n  artifact, report, marker, issue, or repository metadata.\n- Family screen only: never semantically audit every target or build a\n  claim-instrument map. The target-specific commission owns that work.\n- Evidence is not diagnosis: never relabel ordinary friction, a Skill Evolution\n  gate, a conformance defect, calibration pressure, or policy disagreement as a\n  method-gap signal.\n- Safe uncertainty: missing, corrupt, stale, ambiguous, or insufficient evidence\n  never means a skill is healthy or research is not due.\n- No authority escalation: never invoke a method-gap commission, emit a\n  brief-authoring instruction, perform external research, or adopt a prior\n  recommendation.\n- Transient result: do not write a census report. A durable `postpone` or\n  `decline` marker belongs only to a later target-specific census with the\n  owner's consent.\n\n## Procedure\n\n### 1. Resolve and inventory the family\n\nFrom the repository root, run:\n\n```bash\ncargo run --locked -p skill-evidence -- skills method-gap-research-status \\\n  \"\u003Cfamily>\" --root . --now-epoch-milliseconds \"\u003Cclock-ms>\"\n```\n\nChoose one exact census clock in epoch milliseconds before invoking the command;\nreuse that value for any replay. The compiled command validates the selector, enumerates canonical\n`.claude\u002Fskills\u002F` members, hashes their complete packages, validates each\nmatching `reports\u002Fskill-evidence\u002F` event stream through the shared Rust\nskill-evidence authority, derives current-hash use and incident facts in memory,\ndiscovers target-identifying method-gap lineage candidates, extracts\n  bounded hashed headers and lineage-signal lines, and inspects target-scoped\n  Git status and recent commit subjects when Git is available. It writes nothing.\n\nFail safely if the compiled command fails, the shared evidence contract is\nincompatible, or no skill matches. Do not substitute a hand-built family scan\nfor a failed command.\n\n*Done when one complete inventory covers every exact family member and identifies\nevery indeterminate store or unavailable Git observation without changing the\nrepository.*\n\n### 2. Reconstruct only decision-relevant lineage\n\nRead [Lineage reconstruction](references\u002Flineage-reconstruction.md) completely.\nUse each candidate's inventory-provided `header_excerpt`, `signal_lines`, kind,\nsize, and digest first. Read an exact additional section only when those bounded\nprojections leave an ambiguity that could change the member's selection. Never\nread embedded evidence payloads or whole reports merely to prove absence.\nSearch by that target's exact name, path, prior names found in those artifacts,\nand stable report markers only when a named candidate reveals that the inventory\nis incomplete.\n\nFor each family member, establish:\n\n- whether target-specific research is already in flight or returned but\n  unconsumed;\n- the latest audited target identity and any adopted landing;\n- rejected or deferred recommendations and exact reopening triggers;\n- whether later target changes form a stable, observable interval; and\n- whether current evidence is absent, unreadable, on an old hash, or otherwise\n  incapable of supporting a comparison.\n\nIn the census these two count different things: `qualifying_uses` counts runs,\ndeduplicating the several incidents one run can record, while the outcome counts\ncount records. A target whose runs deviated more than once therefore shows\nmore outcomes than uses, and neither number is wrong. Read exercise off the use\ncount and incident volume off the outcome counts; never subtract one from the\nother.\n\nDo not read every target package semantically. A current dirty target is not\nautomatically excluded, but it cannot be selected when its method-bearing\nbaseline is changing too rapidly to freeze coherently.\n\n*Done when every member has one recoverable lineage state, no irrelevant report\nbody or embedded payload was loaded, and every ambiguity that could change\nselection is either resolved from an exact bounded section or explicitly\nindeterminate.*\n\n### 3. Classify candidacy without diagnosing\n\nRead [Selection rules](references\u002Fselection-rules.md) completely. Apply its\nroutes, eligibility gates, ordered signals, and comparison rules to the\ninventory and reconstructed lineage.\n\nReceipts supply interval evidence, not a method-gap verdict. Repeated ordinary\nfriction remains owned by Skill Evolution. A family member becomes a\n`screen next` candidate only when the record supports a plausible\nmissing-capability question with material consequence and a target-specific\ncensus could decide whether external research is due.\n\nAssign every member exactly one status:\n\n- `screen next`;\n- `watch`;\n- `research in flight`;\n- `insufficient evidence`;\n- `no current signal`; or\n- `could not determine`.\n\nThere may be at most one `screen next`. When tied evidence cannot support one\nchoice without arbitrary scoring, assign the tied members `watch` and return no\nselection.\n\n*Done when all members are classified once, ordinary iteration routes remain\nseparate, every positive signal cites concrete current evidence, and at most one\nexact target survives.*\n\n### 4. Render the transient census and stop\n\nReturn:\n\n1. the family selector and number of canonical members;\n2. `Screen next` with the exact target, controlling evidence, material\n   consequence, baseline\u002Flineage state, and why a target-specific census can\n   resolve the uncertainty—or `None`;\n3. the remaining members grouped under their exact statuses with terse reasons;\n4. unavailable or pending evidence;\n5. a literal statement that the census wrote nothing and did not determine that\n   any method gap exists; and\n6. only when one target is selected, this census-only mandate:\n\n```text\nCensus only for \u003Cexact-target-path>. Decide commission now, postpone, or decline; do not author a research brief in this invocation.\n```\n\nNever shorten that mandate to a bare target: a bare target can authorize a\nbrief when the target-specific census recommends `commission now`.\n\n*Done when the output accounts for the whole family, contains zero or one safe\nhandoff, and performs no follow-on action.*\n\n## Completion terminals\n\n- **Selection:** one exact `screen next` target, one census-only mandate, and\n  no mutation or research authorization.\n- **No selection:** every family member classified, the controlling absence,\n  tie, insufficiency, in-flight work, or uncertainty stated, and no invocation.\n- **Safe failure:** the failed contract or instrument named, nothing changed,\n  and no hand-built substitute result.\n","# Method-Gap Research Status\n\nRun one read-only census across an explicitly named family of skills. Recommend\nat most one exact target for a later, census-only method-gap commission.\nThis skill is a portfolio selector, not a method-gap auditor: it never concludes\nthat a method gap exists, decides that research is due, authors a brief, or\ninvokes the commission.\n\nArguments: one skill-name family expressed as a literal name prefix followed by\none terminal `*`, for example `game-*`. Paths, multiple wildcards, character\nclasses, brace expansion, and a bare `*` are refused.\n\n## Hard boundaries\n\n- Read only: never modify a target, evidence stream, stored projection, research\n  artifact, report, marker, issue, or repository metadata.\n- Family screen only: never semantically audit every target or build a\n  claim-instrument map. The target-specific commission owns that work.\n- Evidence is not diagnosis: never relabel ordinary friction, a Skill Evolution\n  gate, a conformance defect, calibration pressure, or policy disagreement as a\n  method-gap signal.\n- Safe uncertainty: missing, corrupt, stale, ambiguous, or insufficient evidence\n  never means a skill is healthy or research is not due.\n- No authority escalation: never invoke a method-gap commission, emit a\n  brief-authoring instruction, perform external research, or adopt a prior\n  recommendation.\n- Transient result: do not write a census report. A durable `postpone` or\n  `decline` marker belongs only to a later target-specific census with the\n  owner's consent.\n\n## Procedure\n\n### 1. Resolve and inventory the family\n\nFrom the repository root, run:\n\n```bash\ncargo run --locked -p skill-evidence -- skills method-gap-research-status \\\n  \"\u003Cfamily>\" --root . --now-epoch-milliseconds \"\u003Cclock-ms>\"\n```\n\nChoose one exact census clock in epoch milliseconds before invoking the command;\nreuse that value for any replay. The compiled command validates the selector, enumerates canonical\n`.claude\u002Fskills\u002F` members, hashes their complete packages, validates each\nmatching `reports\u002Fskill-evidence\u002F` event stream through the shared Rust\nskill-evidence authority, derives current-hash use and incident facts in memory,\ndiscovers target-identifying method-gap lineage candidates, extracts\n  bounded hashed headers and lineage-signal lines, and inspects target-scoped\n  Git status and recent commit subjects when Git is available. It writes nothing.\n\nFail safely if the compiled command fails, the shared evidence contract is\nincompatible, or no skill matches. Do not substitute a hand-built family scan\nfor a failed command.\n\n*Done when one complete inventory covers every exact family member and identifies\nevery indeterminate store or unavailable Git observation without changing the\nrepository.*\n\n### 2. Reconstruct only decision-relevant lineage\n\nRead [Lineage reconstruction](references\u002Flineage-reconstruction.md) completely.\nUse each candidate's inventory-provided `header_excerpt`, `signal_lines`, kind,\nsize, and digest first. Read an exact additional section only when those bounded\nprojections leave an ambiguity that could change the member's selection. Never\nread embedded evidence payloads or whole reports merely to prove absence.\nSearch by that target's exact name, path, prior names found in those artifacts,\nand stable report markers only when a named candidate reveals that the inventory\nis incomplete.\n\nFor each family member, establish:\n\n- whether target-specific research is already in flight or returned but\n  unconsumed;\n- the latest audited target identity and any adopted landing;\n- rejected or deferred recommendations and exact reopening triggers;\n- whether later target changes form a stable, observable interval; and\n- whether current evidence is absent, unreadable, on an old hash, or otherwise\n  incapable of supporting a comparison.\n\nIn the census these two count different things: `qualifying_uses` counts runs,\ndeduplicating the several incidents one run can record, while the outcome counts\ncount records. A target whose runs deviated more than once therefore shows\nmore outcomes than uses, and neither number is wrong. Read exercise off the use\ncount and incident volume off the outcome counts; never subtract one from the\nother.\n\nDo not read every target package semantically. A current dirty target is not\nautomatically excluded, but it cannot be selected when its method-bearing\nbaseline is changing too rapidly to freeze coherently.\n\n*Done when every member has one recoverable lineage state, no irrelevant report\nbody or embedded payload was loaded, and every ambiguity that could change\nselection is either resolved from an exact bounded section or explicitly\nindeterminate.*\n\n### 3. Classify candidacy without diagnosing\n\nRead [Selection rules](references\u002Fselection-rules.md) completely. Apply its\nroutes, eligibility gates, ordered signals, and comparison rules to the\ninventory and reconstructed lineage.\n\nReceipts supply interval evidence, not a method-gap verdict. Repeated ordinary\nfriction remains owned by Skill Evolution. A family member becomes a\n`screen next` candidate only when the record supports a plausible\nmissing-capability question with material consequence and a target-specific\ncensus could decide whether external research is due.\n\nAssign every member exactly one status:\n\n- `screen next`;\n- `watch`;\n- `research in flight`;\n- `insufficient evidence`;\n- `no current signal`; or\n- `could not determine`.\n\nThere may be at most one `screen next`. When tied evidence cannot support one\nchoice without arbitrary scoring, assign the tied members `watch` and return no\nselection.\n\n*Done when all members are classified once, ordinary iteration routes remain\nseparate, every positive signal cites concrete current evidence, and at most one\nexact target survives.*\n\n### 4. Render the transient census and stop\n\nReturn:\n\n1. the family selector and number of canonical members;\n2. `Screen next` with the exact target, controlling evidence, material\n   consequence, baseline\u002Flineage state, and why a target-specific census can\n   resolve the uncertainty—or `None`;\n3. the remaining members grouped under their exact statuses with terse reasons;\n4. unavailable or pending evidence;\n5. a literal statement that the census wrote nothing and did not determine that\n   any method gap exists; and\n6. only when one target is selected, this census-only mandate:\n\n```text\nCensus only for \u003Cexact-target-path>. Decide commission now, postpone, or decline; do not author a research brief in this invocation.\n```\n\nNever shorten that mandate to a bare target: a bare target can authorize a\nbrief when the target-specific census recommends `commission now`.\n\n*Done when the output accounts for the whole family, contains zero or one safe\nhandoff, and performs no follow-on action.*\n\n## Completion terminals\n\n- **Selection:** one exact `screen next` target, one census-only mandate, and\n  no mutation or research authorization.\n- **No selection:** every family member classified, the controlling absence,\n  tie, insufficiency, in-flight work, or uncertainty stated, and no invocation.\n- **Safe failure:** the failed contract or instrument named, nothing changed,\n  and no hand-built substitute result.\n",{"name":5,"description":7,"argument-hint":23,"disable-model-invocation":24},"\u003Cskill-name-family, for example game-*>",true,"main","9b13db1df193bd0709752040255bd64b682d3242",".claude\u002Fskills\u002Fmethod-gap-research-status","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence","https:\u002F\u002Fdocs.rs\u002Fskill-evidence","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck","2026-08-23 17:23:53",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":24},[34,37,40],{"tool":35,"command":36},"npx","npx skills add joeloverbeck\u002Fskill-evidence --skill method-gap-research-status",{"tool":38,"command":39},"bunx","bunx skills add joeloverbeck\u002Fskill-evidence --skill method-gap-research-status",{"tool":41,"command":42},"pnpm","pnpm dlx skills add joeloverbeck\u002Fskill-evidence --skill method-gap-research-status","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmethod-gap-research-status\u002FSKILL.md",[45,51,55],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"agents\u002Fopenai.yaml","REFERENCE","参考文档",349,"https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmethod-gap-research-status\u002Fagents\u002Fopenai.yaml",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"references\u002Flineage-reconstruction.md",2215,"https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmethod-gap-research-status\u002Freferences\u002Flineage-reconstruction.md",{"path":56,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":57,"sourceHtmlUrl":58},"references\u002Fselection-rules.md",6026,"https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmethod-gap-research-status\u002Freferences\u002Fselection-rules.md",[60,67,72,79,84,89],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":18},"01M7AAQDXB1QQET7VSTF1FCEET","triage","Move issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftriage\u002Freviews\u002Ffce6031a-a8f1-44b9-93cc-5c27488a51d3\u002Fcandidate\u002FSKILL.md",[],[],{"skillNo":68,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":18},"01M7AAQBCSHN0ZH9MHQZNEHCGR","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftriage\u002Freviews\u002F48b04150-c9c8-42a7-982a-324aceb6d21e\u002Fheld\u002Fember\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":18},"01M7AAQ8BDCGNPBVJVGQ1KWSN7","tdd","Test-driven development. Use when the user wants a red-green development loop or integration tests.","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftdd\u002Freviews\u002Frev_c25c6d82-756b-49e4-85ed-bd164820f2f2\u002Fpre-land-backup\u002FSKILL.md",[],[],{"skillNo":80,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":18},"01M7AAQ6DRHY95HPCDGG7TBP93","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftdd\u002Freviews\u002Frev_c25c6d82-756b-49e4-85ed-bd164820f2f2\u002Fcandidate\u002FSKILL.md",[],[],{"skillNo":85,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":18},"01M7AAQ3XN1HSP13RC35NFTKT0","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftdd\u002Freviews\u002Frev_3e39da1f-a737-4029-bd88-f710519fe3de\u002Fpre-land-backup\u002FSKILL.md",[],[],{"skillNo":90,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":18},"01M7AAQ1NGVGK62GB5Z6JSNJ76","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fskill-evidence\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftdd\u002Freviews\u002Frev_3e39da1f-a737-4029-bd88-f710519fe3de\u002Fcandidate\u002FSKILL.md",[],[]]