[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M71J3C44W2CNDX4Q3499NHTG":3,"skill-same:01M71J3C44W2CNDX4Q3499NHTG":57},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":26,"linkOnly":27,"markdownContent":28,"markdownBody":29,"frontMatter":30,"refName":31,"commitSha":32,"skillDirectory":33,"forksCount":13,"repositoryHtmlUrl":34,"repositoryHomepageUrl":6,"ownerHtmlUrl":35,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":36,"install":37,"resources":50},"01M71J3C44W2CNDX4Q3499NHTG","implement",null,"Creates a branch, plans the implementation, and makes code changes based on an Issue or an instruction. Accepts an Issue number or a text instruction.","ozzy-labs\u002Fskills","ozzy-labs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F246109093?v=4","01M6H0M2K2PZXHWNHJJTWY20FY","01M71J2B6FZ5KPEXPR5VS1TX7Z",0,"MIT","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fimplement\u002FSKILL.md",[],[5,18,19,20,21,22,23,24,25],"text","branch","changes","plans","issue","implementation","number","code","2026-11-10 10:12:30",false,"---\nname: implement\ndescription: Creates a branch, plans the implementation, and makes code changes based on an Issue or an instruction. Accepts an Issue number or a text instruction.\n---\n\n# implement - Create a branch and implement from an Issue\u002Finstruction\n\nBased on reading an Issue or a direct instruction, this performs everything from branch creation through implementation planning to code changes.\n\n## Input\n\n- **If an Issue number** (`#N` or a bare number): fetch the content with `gh issue view \u003CN>` and organize the requirements\n- **If a text instruction**: treat it as-is as the requirements\n- **If no argument**: confirm what to implement\n\n## Action classification and policy reference\n\nThis skill does not hardcode individual approval gates into prose; instead it follows the central autonomy policy (the SSOT for the 3 classes, gate vocabulary, and `policy.yaml` hierarchy defined by the `policy` skill). It classifies its own actions into the following classes, resolves the effective gate, and then executes:\n\n| Action in this skill | Class | policy reference | Zero-config default gate |\n| --- | --- | --- | --- |\n| Implementation on the branch (file edit\u002Fadd\u002Fsafe delete) | `reversible-local` | `--action=branch-edit` | `proceed` (present the plan and continue + audit trail) |\n| Changes involving migration \u002F data deletion \u002F CI or release config changes | `irreversible` | `--class=irreversible` | `ask` (explicit approval before starting) |\n\nResolve the effective gate with `policy-read.mjs` in the sibling `policy` skill (for Claude Code user-scope this is `~\u002F.claude\u002Fskills\u002Fpolicy\u002Fpolicy-read.mjs`, for dogfooding `\u003Crepo>\u002F.claude\u002Fskills\u002Fpolicy\u002Fpolicy-read.mjs`, and for Codex\u002FGemini `.agents\u002Fskills\u002Fpolicy\u002Fpolicy-read.mjs`):\n\n```bash\nnode \u003Cpolicy skill のディレクトリ>\u002Fpolicy-read.mjs --action=branch-edit --repo-root=\"$PWD\"\n# => .resolved.gate（既定 proceed）\nnode \u003Cpolicy skill のディレクトリ>\u002Fpolicy-read.mjs --class=irreversible --repo-root=\"$PWD\"\n# => .resolved.gate（既定 ask）\n```\n\nThe gate vocabulary has only 3 values:\n\n- `proceed`: execute without waiting for approval, and leave the plan\u002Fchanges in the report as an audit trail\n- `batch-confirm`: confirm once, all together, before starting\n- `ask`: obtain explicit approval (Approval Gate) for each action\n\n**Does not break even if policy is absent:** `policy-read.mjs` is fail-safe by design — unreadable or invalid values always fall back to the stricter side (`ask`). In environments where the `policy` skill itself is not deployed and `policy-read.mjs` cannot be called, apply the zero-config default gates from the table above directly (`reversible-local`=`proceed` \u002F `irreversible`=`ask`).\n\n**Consistency under drive:** drive is delegated autonomous execution by the user, and the `reversible-local` default of `proceed` (skip plan approval and continue) is consistent with this delegation. Standard branch implementation proceeds with proceed, and only changes judged `irreversible` require approval with gate=`ask`.\n\n## Procedure\n\n### 1. Create a branch\n\n1. Check the current state with `git status` and `git branch --show-current`\n2. Determine a branch name in `\u003Ctype>\u002F\u003Cslug>` format from the requirements\n3. Create the branch with `git checkout -b \u003Cbranch-name>`\n\nIf already on a feature branch, confirm whether to continue working on that branch.\n\n### 2. Implementation plan and action classification\n\n1. Investigate the codebase\n   - Identify relevant files\n   - Understand existing implementation patterns\n   - Check the scope of impact\n2. Present the implementation plan:\n   - Files to change and their content\n   - Scope of impact\n3. Classify the change into a class using the table in \"Action classification and policy reference,\" and resolve the effective policy gate:\n   - Only ordinary implementation on the branch → `reversible-local` (default `proceed`): present the plan and continue (do not wait for approval; the plan is kept as an audit trail)\n   - Includes migration \u002F data deletion \u002F CI or release config changes → `irreversible` (default `ask`): obtain explicit approval before proceeding to implementation\n\n### 3. Implementation\n\nExecute code changes according to the gate resolved by policy (`proceed` starts without waiting for approval, `batch-confirm` confirms all at once before starting, `ask` starts after approval). Report progress as each file's changes are completed.\n\n### 4. Verification (verify)\n\nAfter implementation is complete, refer to `.agents\u002Fskills\u002Fverify\u002FSKILL.md` and run the combined verification (build + type + test + lint) with the verify engine. verify auto-discovers verification commands through a discovery chain (AGENTS.md's 「検証」 section → package.json scripts → task runner → language heuristics) and runs them serially, each with its source attributed.\n\nIf errors occur, fix them on the spot and run verify again.\n\n### 5. Completion report\n\n```text\n実装完了:\n  ブランチ: \u003Cbranch-name>\n  変更ファイル:\n    A path\u002Fto\u002Fnew-file\n    M path\u002Fto\u002Fmodified-file\n```\n\n## Notes\n\n- Do not read or stage `.env` files\n- If the `gh` CLI is not authenticated, display an error message and abort\n- Classify actions into the 3 classes and resolve policy before executing. Do not hardcode individual approval gates into prose\n- For changes judged `irreversible` (migration \u002F data deletion \u002F CI or release config changes), always obtain explicit approval under gate=`ask` before starting\n","# implement - Create a branch and implement from an Issue\u002Finstruction\n\nBased on reading an Issue or a direct instruction, this performs everything from branch creation through implementation planning to code changes.\n\n## Input\n\n- **If an Issue number** (`#N` or a bare number): fetch the content with `gh issue view \u003CN>` and organize the requirements\n- **If a text instruction**: treat it as-is as the requirements\n- **If no argument**: confirm what to implement\n\n## Action classification and policy reference\n\nThis skill does not hardcode individual approval gates into prose; instead it follows the central autonomy policy (the SSOT for the 3 classes, gate vocabulary, and `policy.yaml` hierarchy defined by the `policy` skill). It classifies its own actions into the following classes, resolves the effective gate, and then executes:\n\n| Action in this skill | Class | policy reference | Zero-config default gate |\n| --- | --- | --- | --- |\n| Implementation on the branch (file edit\u002Fadd\u002Fsafe delete) | `reversible-local` | `--action=branch-edit` | `proceed` (present the plan and continue + audit trail) |\n| Changes involving migration \u002F data deletion \u002F CI or release config changes | `irreversible` | `--class=irreversible` | `ask` (explicit approval before starting) |\n\nResolve the effective gate with `policy-read.mjs` in the sibling `policy` skill (for Claude Code user-scope this is `~\u002F.claude\u002Fskills\u002Fpolicy\u002Fpolicy-read.mjs`, for dogfooding `\u003Crepo>\u002F.claude\u002Fskills\u002Fpolicy\u002Fpolicy-read.mjs`, and for Codex\u002FGemini `.agents\u002Fskills\u002Fpolicy\u002Fpolicy-read.mjs`):\n\n```bash\nnode \u003Cpolicy skill のディレクトリ>\u002Fpolicy-read.mjs --action=branch-edit --repo-root=\"$PWD\"\n# => .resolved.gate（既定 proceed）\nnode \u003Cpolicy skill のディレクトリ>\u002Fpolicy-read.mjs --class=irreversible --repo-root=\"$PWD\"\n# => .resolved.gate（既定 ask）\n```\n\nThe gate vocabulary has only 3 values:\n\n- `proceed`: execute without waiting for approval, and leave the plan\u002Fchanges in the report as an audit trail\n- `batch-confirm`: confirm once, all together, before starting\n- `ask`: obtain explicit approval (Approval Gate) for each action\n\n**Does not break even if policy is absent:** `policy-read.mjs` is fail-safe by design — unreadable or invalid values always fall back to the stricter side (`ask`). In environments where the `policy` skill itself is not deployed and `policy-read.mjs` cannot be called, apply the zero-config default gates from the table above directly (`reversible-local`=`proceed` \u002F `irreversible`=`ask`).\n\n**Consistency under drive:** drive is delegated autonomous execution by the user, and the `reversible-local` default of `proceed` (skip plan approval and continue) is consistent with this delegation. Standard branch implementation proceeds with proceed, and only changes judged `irreversible` require approval with gate=`ask`.\n\n## Procedure\n\n### 1. Create a branch\n\n1. Check the current state with `git status` and `git branch --show-current`\n2. Determine a branch name in `\u003Ctype>\u002F\u003Cslug>` format from the requirements\n3. Create the branch with `git checkout -b \u003Cbranch-name>`\n\nIf already on a feature branch, confirm whether to continue working on that branch.\n\n### 2. Implementation plan and action classification\n\n1. Investigate the codebase\n   - Identify relevant files\n   - Understand existing implementation patterns\n   - Check the scope of impact\n2. Present the implementation plan:\n   - Files to change and their content\n   - Scope of impact\n3. Classify the change into a class using the table in \"Action classification and policy reference,\" and resolve the effective policy gate:\n   - Only ordinary implementation on the branch → `reversible-local` (default `proceed`): present the plan and continue (do not wait for approval; the plan is kept as an audit trail)\n   - Includes migration \u002F data deletion \u002F CI or release config changes → `irreversible` (default `ask`): obtain explicit approval before proceeding to implementation\n\n### 3. Implementation\n\nExecute code changes according to the gate resolved by policy (`proceed` starts without waiting for approval, `batch-confirm` confirms all at once before starting, `ask` starts after approval). Report progress as each file's changes are completed.\n\n### 4. Verification (verify)\n\nAfter implementation is complete, refer to `.agents\u002Fskills\u002Fverify\u002FSKILL.md` and run the combined verification (build + type + test + lint) with the verify engine. verify auto-discovers verification commands through a discovery chain (AGENTS.md's 「検証」 section → package.json scripts → task runner → language heuristics) and runs them serially, each with its source attributed.\n\nIf errors occur, fix them on the spot and run verify again.\n\n### 5. Completion report\n\n```text\n実装完了:\n  ブランチ: \u003Cbranch-name>\n  変更ファイル:\n    A path\u002Fto\u002Fnew-file\n    M path\u002Fto\u002Fmodified-file\n```\n\n## Notes\n\n- Do not read or stage `.env` files\n- If the `gh` CLI is not authenticated, display an error message and abort\n- Classify actions into the 3 classes and resolve policy before executing. Do not hardcode individual approval gates into prose\n- For changes judged `irreversible` (migration \u002F data deletion \u002F CI or release config changes), always obtain explicit approval under gate=`ask` before starting\n",{"name":5,"description":7},"main","db212a47efe751fa1c42bd76e61dbfd8328e5b10",".agents\u002Fskills\u002Fimplement","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills","https:\u002F\u002Fgithub.com\u002Fozzy-labs","2026-07-13 06:49:42",{"source":8,"skillName":5,"commands":38,"prompt":48,"zipAvailable":49},[39,42,45],{"tool":40,"command":41},"npx","npx skills add ozzy-labs\u002Fskills --skill implement",{"tool":43,"command":44},"bunx","bunx skills add ozzy-labs\u002Fskills --skill implement",{"tool":46,"command":47},"pnpm","pnpm dlx skills add ozzy-labs\u002Fskills --skill implement","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fimplement\u002FSKILL.md",true,[51],{"path":52,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":55,"sourceHtmlUrl":56},"SKILL.claude-code.md","REFERENCE","参考文档",1908,"https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fimplement\u002FSKILL.claude-code.md",[58,84,109,131,163,185],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":26},"01M71J6R63H4E6ZE6KMDM4CT85","verify","An integrated skill that runs the combined validation of build \u002F typecheck \u002F test \u002F lint in one shot. The `verify.mjs` engine auto-discovers validation commands via a discovery chain (AGENTS.md's 「検証」 section → package.json scripts → justfile\u002FMakefile\u002Flefthook → language heuristics), executes them serially with their source attached, and returns a result summary. If a higher-priority stage produces a hit, only that stage is executed (no crossing stages).","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fverify\u002FSKILL.md",[],[60,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83],"combined","heuristics","chain","repo","discovery","section","lefthook","only","auto","package","stage","Bash","one","lint","dry","engine","with","test","typecheck",{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":26},"01M71J6JW7SAY9PDKAWHG3SJCR","usage-guard","Monitors Claude Code's Usage Limit (5-hour = Current \u002F weekly = Weekly), pauses work when it exceeds 95%, and automatically resumes once the reset restores headroom. Bundles an engine form that a caller such as drive Reads at a checkpoint, and a standalone form, `\u002Fusage-guard \"\u003Ccontinuation command>\"`, that guards arbitrary work. Claude-only (depends on the OAuth usage-rate endpoint + ScheduleWakeup).","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fusage-guard\u002FSKILL.md",[],[91,92,93,94,95,72,96,97,25,98,99,100,101,102,103,104,105,106,107,108],"usage","usage guard","guard","weekly","standalone","guards","when","reset","work","command","at","rate","status","endpoint","as","continuation","form","limit",{"skillNo":110,"name":111,"nameZh":6,"description":112,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":113,"categories":114,"keywords":115,"publishedAt":26},"01M71J6FRG119REA1G0S99Q4GZ","topics","Selects GitHub topic candidates by having the `topics.mjs` engine perform constraint validation, popularity measurement, broad+narrow \u002F singular-plural comparison, and hardcoded ozzy-labs conventions, then applies them via `gh repo edit --add-topic` following the policy's `externally-visible` gate (default batch-confirm). Scope is limited to use within ozzy-labs.","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Ftopics\u002FSKILL.md",[],[111,116,117,68,118,119,120,121,122,123,124,76,125,126,127,79,80,128,129,130],"measurement","scope","policy","default","true","gh","plural","use","comparison","apply","candidate","batch","candidates","github","via",{"skillNo":132,"name":133,"nameZh":6,"description":134,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":135,"categories":136,"keywords":147,"publishedAt":26},"01M71J6DC27P4FDSP10SDT2PSN","skill-observability","Defines the skill observability event contract (event.schema.json) and emit substrate (obs-emit.mjs). The SSOT for the measurement layer of the skill improvement loop. Referenced by other skills and hooks.","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fskill-observability\u002FSKILL.md",[137,142],{"categoryNo":138,"categoryName":139,"taxonomyType":140,"taxonomyTypeDesc":141},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",{"categoryNo":143,"categoryName":144,"taxonomyType":145,"taxonomyTypeDesc":146},"01JBSEEDCAT000000000000037","律师与法务","PROFESSION","适用职业",[148,149,116,150,151,152,153,154,155,156,157,158,159,160,161,162],"observability","skill observability","improvement","skill improvement","for","event","hooks","contract","schema","loop","false","substrate","json","improvement loop","layer",{"skillNo":164,"name":165,"nameZh":6,"description":166,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":167,"categories":168,"keywords":170,"publishedAt":26},"01M71J6AYECH1J94RAZ49QZ19Y","skill-metrics","Aggregates the local observability event log (~\u002F.agents\u002Fobservability\u002Fevents.jsonl) and presents, read-only, per-skill invocation counts and notable events (fallback \u002F HITL rejection \u002F loop cap reached \u002F abort). Fires on \"show me the skill metrics,\" \"aggregate the observability results,\" \"which skills get used the most?\" Does not send anything.","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fskill-metrics\u002FSKILL.md",[169],{"categoryNo":138,"categoryName":139,"taxonomyType":140,"taxonomyTypeDesc":141},[171,172,148,173,174,175,176,177,153,178,179,180,181,120,72,157,182,76,183,184],"skill metrics","metrics","event log","which","show","get","read-only","iso","aggregates","send","local","show me","log","jsonl",{"skillNo":186,"name":187,"nameZh":6,"description":188,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":189,"categories":190,"keywords":191,"publishedAt":26},"01M71J68TD0QFATGW3BBZVCWA2","ship","Runs verification, commit, and PR creation all at once. An integrated pipeline that runs verify (build + type + test + lint) → commit → PR creation in sequence against the changes.","https:\u002F\u002Fgithub.com\u002Fozzy-labs\u002Fskills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fship\u002FSKILL.md",[],[187,192,193,82,20,194,60,195,196,197,101,198,199,200,120,201,76,202,78],"type","verification","pr","creation","build","pipeline","read","sequence","grep","pr creation","commit"]