[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6CZGG2GJZ7PGVTZSG8NDB2R":3,"skill-same:01M6CZGG2GJZ7PGVTZSG8NDB2R":71},{"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":22,"publishedAt":43,"linkOnly":44,"markdownContent":45,"markdownBody":46,"frontMatter":47,"refName":49,"commitSha":50,"skillDirectory":51,"forksCount":13,"repositoryHtmlUrl":52,"repositoryHomepageUrl":53,"ownerHtmlUrl":54,"ownerBlogUrl":55,"ownerDescription":6,"updatedAt":56,"install":57,"resources":70},"01M6CZGG2GJZ7PGVTZSG8NDB2R","run-task",null,"Runs a backlog item (bug, feature, task, epic, milestone, phase) end-to-end by gathering context, constructing a startup prompt, spawning a worktree-backed subagent, and managing the claim\u002Fdone\u002Freview cycle. Use when the user wants to execute a backlog target like B123, F456, P2.M4.E3, etc.","clankercode\u002Fclawq","clankercode","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F269201082?v=4","01M6CZFKG57M58N3F63P2H9742","01M6CZFKMHDHR7D6ADH22V9T11",0,"Unlicense","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002Fdocs\u002Fbuiltin-skills\u002Frun-task\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42],"run","run task","task","worktree","review","feature","backlog","review cycle","done","claim","milestone","cycle","startup","bug","use","execute","user","prompt","extra","subagent","2026-11-10 10:12:30",false,"---\nname: run-task\ndescription: Runs a backlog item (bug, feature, task, epic, milestone, phase) end-to-end by gathering context, constructing a startup prompt, spawning a worktree-backed subagent, and managing the claim\u002Fdone\u002Freview cycle. Use when the user wants to execute a backlog target like B123, F456, P2.M4.E3, etc.\nargument-hint: \u003CTARGET> [extra steering]\n---\n\n# run-task\n\nExecute backlog target `$ARGUMENTS` end-to-end: gather context, build a startup prompt, spawn a native\u002Flocal or worktree-backed subagent, and manage the claim\u002Fdone\u002Freview cycle.\n\n## Setup\n\nGet backlog usage reference:\n!`bl howto`\n\n## Step 1: Classify the target\n\nInterpret `$ARGUMENTS` as a backlog target. If none provided, ask for one.\n\n| Class | Examples |\n|-------|----------|\n| Leaf item | `B123`, `F123`, `I123`, `P1.M2.E3.T4` |\n| Scope item | epic `P1.M2.E3`, milestone `P1.M2`, phase `P1` |\n\n## Step 2: Gather backlog context\n\n### Leaf items\nRun `bl claim TARGET` — capture the full output. If claim fails due to existing ownership, surface that explicitly and decide whether to continue, hand off, or reuse an existing worker.\n\n### Scope items\nRun all three:\n- `bl show TARGET`\n- `bl tree TARGET --unfinished`\n- `bl list TARGET --available --json`\n\nSummarize runnable epics\u002Ftasks from the available list.\n\n## Step 3: Construct the startup prompt\n\nBuild the prompt from the command outputs gathered above. If the user gave extra steering, append it after a final `---` separator.\n\n### Leaf prompt\n\n```text\n$ bl howto\n\u003COUTPUT OF bl howto>\n\n---\n\n$ bl claim TARGET\n\u003COUTPUT OF bl claim TARGET>\n\n---\n\n\u003CTAIL INSTRUCTION>\n```\n\n**Leaf tail (normal):**\n> Complete TARGET, then mark it done, then load and use your `review-and-fix` skill.\n\n**Leaf tail (plan mode):**\n> Plan and complete TARGET. After planning, run `git rebase` against your parent branch to ensure you're up to date. Make `bl done TARGET` the second-to-last step of the plan, and make loading and using the `review-and-fix` skill the final step of the plan.\n\nUse plain Git for parent-branch checks and rebases. Do not use Graphite (`gt`)\nor other stack-management tools to infer or mutate branch relationships.\n\n### Scope prompt (epic, milestone, phase)\n\n```text\n$ bl howto\n\u003COUTPUT OF bl howto>\n\n---\n\n$ bl show TARGET\n\u003COUTPUT OF bl show TARGET>\n\n---\n\n$ bl tree TARGET --unfinished\n\u003COUTPUT OF bl tree TARGET --unfinished>\n\n---\n\n$ bl list TARGET --available --json\n\u003COUTPUT OF bl list TARGET --available --json>\n\n---\n\nRunnable epics now:\n\u003CRUNNABLE SUMMARY>\n\n---\n\n\u003CTAIL INSTRUCTION>\n```\n\n**Epic tail:**\n> Complete epic TARGET. Use `bl list TARGET --available --json` to identify runnable tasks in this epic. Claim the relevant runnable tasks for TARGET, complete them, mark them done, and refresh epic availability until the epic is complete or blocked. When the epic is complete, load and use `review-and-fix`, then commit. If all remaining work is blocked, summarize blockers and stop.\n\n**Milestone\u002FPhase tail:**\n> Work through TARGET epic-by-epic. Use the scoped tree and runnable epic summary to choose an epic in this scope that currently has runnable tasks. Complete one epic before moving to the next. Within each epic, claim the runnable tasks, complete them, mark them done, and refresh epic availability until that epic is complete or blocked. After each completed epic, load and use `review-and-fix`, then commit. Refresh the scope state and continue until no runnable epics remain. If blocked, summarize the blockers and stop.\n\n## Step 4: Choose runner and agent role\n\nPrefer the local runner for most tasks. Escalate to external only when genuinely needed.\n\n| Sizing | Runner | When |\n|--------|--------|------|\n| Small\u002Fmedium | `local` + agent role | Default. Most leaf items and single-epic scope. |\n| Large (>8h, cross-cutting) | External (`opencode`, `claude`) | Multi-file refactors spanning the whole codebase. |\n| Entire project | `local` team-lead + local coders | Orchestrator spawns per-epic\u002Ftask agents. |\n\n### Agent roles (local runner)\n\n| Task type | agent_name | use_worktree |\n|-----------|------------|--------------|\n| Implement feature\u002Ffix | `coder` | true |\n| Write\u002Frun tests | `tester` | true |\n| Refactor\u002Fcleanup | `refactorer` | true |\n| Debug\u002Froot-cause | `debugger` | true |\n| Review code | `reviewer` | false |\n| Explore\u002Fresearch | `researcher` | false |\n| Plan architecture | `planner` | false |\n| Orchestrate subtasks | `team-lead` | false |\n| Run entire project | `ceo` | false |\n\n## Step 5: Launch the subagent\n\n```\nbackground_task_enqueue(\n  runner: \"local\",\n  agent_name: \"\u003Crole>\",\n  repo_path: \"\u003Crepo root>\",\n  prompt: \"\u003Cconstructed prompt>\",\n  branch: \"clawq-bg-\u003CTARGET>\",\n  use_worktree: true\u002Ffalse\n)\n```\n\nAutomerge is enabled by default. To disable it for a specific task, pass `automerge: false`.\n\n## Step 6: Monitor and steer\n\n- `background_task_list` — check status\n- `background_task_logs` \u002F `background_task_wait` — follow progress\n- `background_task_transcript` — inspect bounded conversation history, with regex filters and JSONL export for large results\n- `background_task_send_message` — send clarifications\n- `background_task_resume` \u002F `background_task_recover` — handle stalls\n\nVerify the worker is making progress; don't assume success from a clean launch.\n\n## Step 7: Close out\n\n- When a worktree-backed task finishes, the system automatically sends a **completion pass** message to the agent. The agent resumes with its session context, commits remaining changes, rebases against master, reviews all changes for correctness\u002Fquality\u002Fcompleteness\u002Fsafety (review-and-fix guard), runs checks, and outputs the sentinel `OK_TASK_DONE_CHECKED_REBASED_COMMITED`.\n- Since automerge is enabled by default, the system then attempts a fast-forward merge. If automerge was disabled, the user is notified normally.\n- Inspect changes and verify tests\u002Freview status.\n- Use `background_finalize(id=...)` only when manual rebase and fast-forward merge is desired after the completion pass.\n- If review reveals follow-up issues, defer done state until after review\u002Ffix is clean.\n\n## Constraints\n\n- Reconstruct prompts from backlog outputs.\n- Surface claim conflicts explicitly.\n- Use clawq-native worktree\u002Fbackground tools over ad-hoc shell workflows.\n- For interactive task completion, use plain Git for branch ancestry and rebase\n  steps; do not call Graphite (`gt`) unless the user explicitly asks for it.\n- Branch names should visibly contain the target ID (e.g. `clawq-bg-B467`).\n","# run-task\n\nExecute backlog target `$ARGUMENTS` end-to-end: gather context, build a startup prompt, spawn a native\u002Flocal or worktree-backed subagent, and manage the claim\u002Fdone\u002Freview cycle.\n\n## Setup\n\nGet backlog usage reference:\n!`bl howto`\n\n## Step 1: Classify the target\n\nInterpret `$ARGUMENTS` as a backlog target. If none provided, ask for one.\n\n| Class | Examples |\n|-------|----------|\n| Leaf item | `B123`, `F123`, `I123`, `P1.M2.E3.T4` |\n| Scope item | epic `P1.M2.E3`, milestone `P1.M2`, phase `P1` |\n\n## Step 2: Gather backlog context\n\n### Leaf items\nRun `bl claim TARGET` — capture the full output. If claim fails due to existing ownership, surface that explicitly and decide whether to continue, hand off, or reuse an existing worker.\n\n### Scope items\nRun all three:\n- `bl show TARGET`\n- `bl tree TARGET --unfinished`\n- `bl list TARGET --available --json`\n\nSummarize runnable epics\u002Ftasks from the available list.\n\n## Step 3: Construct the startup prompt\n\nBuild the prompt from the command outputs gathered above. If the user gave extra steering, append it after a final `---` separator.\n\n### Leaf prompt\n\n```text\n$ bl howto\n\u003COUTPUT OF bl howto>\n\n---\n\n$ bl claim TARGET\n\u003COUTPUT OF bl claim TARGET>\n\n---\n\n\u003CTAIL INSTRUCTION>\n```\n\n**Leaf tail (normal):**\n> Complete TARGET, then mark it done, then load and use your `review-and-fix` skill.\n\n**Leaf tail (plan mode):**\n> Plan and complete TARGET. After planning, run `git rebase` against your parent branch to ensure you're up to date. Make `bl done TARGET` the second-to-last step of the plan, and make loading and using the `review-and-fix` skill the final step of the plan.\n\nUse plain Git for parent-branch checks and rebases. Do not use Graphite (`gt`)\nor other stack-management tools to infer or mutate branch relationships.\n\n### Scope prompt (epic, milestone, phase)\n\n```text\n$ bl howto\n\u003COUTPUT OF bl howto>\n\n---\n\n$ bl show TARGET\n\u003COUTPUT OF bl show TARGET>\n\n---\n\n$ bl tree TARGET --unfinished\n\u003COUTPUT OF bl tree TARGET --unfinished>\n\n---\n\n$ bl list TARGET --available --json\n\u003COUTPUT OF bl list TARGET --available --json>\n\n---\n\nRunnable epics now:\n\u003CRUNNABLE SUMMARY>\n\n---\n\n\u003CTAIL INSTRUCTION>\n```\n\n**Epic tail:**\n> Complete epic TARGET. Use `bl list TARGET --available --json` to identify runnable tasks in this epic. Claim the relevant runnable tasks for TARGET, complete them, mark them done, and refresh epic availability until the epic is complete or blocked. When the epic is complete, load and use `review-and-fix`, then commit. If all remaining work is blocked, summarize blockers and stop.\n\n**Milestone\u002FPhase tail:**\n> Work through TARGET epic-by-epic. Use the scoped tree and runnable epic summary to choose an epic in this scope that currently has runnable tasks. Complete one epic before moving to the next. Within each epic, claim the runnable tasks, complete them, mark them done, and refresh epic availability until that epic is complete or blocked. After each completed epic, load and use `review-and-fix`, then commit. Refresh the scope state and continue until no runnable epics remain. If blocked, summarize the blockers and stop.\n\n## Step 4: Choose runner and agent role\n\nPrefer the local runner for most tasks. Escalate to external only when genuinely needed.\n\n| Sizing | Runner | When |\n|--------|--------|------|\n| Small\u002Fmedium | `local` + agent role | Default. Most leaf items and single-epic scope. |\n| Large (>8h, cross-cutting) | External (`opencode`, `claude`) | Multi-file refactors spanning the whole codebase. |\n| Entire project | `local` team-lead + local coders | Orchestrator spawns per-epic\u002Ftask agents. |\n\n### Agent roles (local runner)\n\n| Task type | agent_name | use_worktree |\n|-----------|------------|--------------|\n| Implement feature\u002Ffix | `coder` | true |\n| Write\u002Frun tests | `tester` | true |\n| Refactor\u002Fcleanup | `refactorer` | true |\n| Debug\u002Froot-cause | `debugger` | true |\n| Review code | `reviewer` | false |\n| Explore\u002Fresearch | `researcher` | false |\n| Plan architecture | `planner` | false |\n| Orchestrate subtasks | `team-lead` | false |\n| Run entire project | `ceo` | false |\n\n## Step 5: Launch the subagent\n\n```\nbackground_task_enqueue(\n  runner: \"local\",\n  agent_name: \"\u003Crole>\",\n  repo_path: \"\u003Crepo root>\",\n  prompt: \"\u003Cconstructed prompt>\",\n  branch: \"clawq-bg-\u003CTARGET>\",\n  use_worktree: true\u002Ffalse\n)\n```\n\nAutomerge is enabled by default. To disable it for a specific task, pass `automerge: false`.\n\n## Step 6: Monitor and steer\n\n- `background_task_list` — check status\n- `background_task_logs` \u002F `background_task_wait` — follow progress\n- `background_task_transcript` — inspect bounded conversation history, with regex filters and JSONL export for large results\n- `background_task_send_message` — send clarifications\n- `background_task_resume` \u002F `background_task_recover` — handle stalls\n\nVerify the worker is making progress; don't assume success from a clean launch.\n\n## Step 7: Close out\n\n- When a worktree-backed task finishes, the system automatically sends a **completion pass** message to the agent. The agent resumes with its session context, commits remaining changes, rebases against master, reviews all changes for correctness\u002Fquality\u002Fcompleteness\u002Fsafety (review-and-fix guard), runs checks, and outputs the sentinel `OK_TASK_DONE_CHECKED_REBASED_COMMITED`.\n- Since automerge is enabled by default, the system then attempts a fast-forward merge. If automerge was disabled, the user is notified normally.\n- Inspect changes and verify tests\u002Freview status.\n- Use `background_finalize(id=...)` only when manual rebase and fast-forward merge is desired after the completion pass.\n- If review reveals follow-up issues, defer done state until after review\u002Ffix is clean.\n\n## Constraints\n\n- Reconstruct prompts from backlog outputs.\n- Surface claim conflicts explicitly.\n- Use clawq-native worktree\u002Fbackground tools over ad-hoc shell workflows.\n- For interactive task completion, use plain Git for branch ancestry and rebase\n  steps; do not call Graphite (`gt`) unless the user explicitly asks for it.\n- Branch names should visibly contain the target ID (e.g. `clawq-bg-B467`).\n",{"name":5,"description":7,"argument-hint":48},"\u003CTARGET> [extra steering]","master","7c87c0cf8d869c392efbe4b2f0065c85657bf969","docs\u002Fbuiltin-skills\u002Frun-task","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq","https:\u002F\u002Fclawq.org\u002F","https:\u002F\u002Fgithub.com\u002Fclankercode","xk.io","2026-11-12 11:53:34",{"source":8,"skillName":5,"commands":58,"prompt":68,"zipAvailable":69},[59,62,65],{"tool":60,"command":61},"npx","npx skills add clankercode\u002Fclawq --skill run-task",{"tool":63,"command":64},"bunx","bunx skills add clankercode\u002Fclawq --skill run-task",{"tool":66,"command":67},"pnpm","pnpm dlx skills add clankercode\u002Fclawq --skill run-task","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002Fdocs\u002Fbuiltin-skills\u002Frun-task\u002FSKILL.md",true,[],[72,101,111,133,144,155],{"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":81,"publishedAt":43},"01M6CZGD2YWGXTV25FM004F7ZM","pipeline-designer","Design and create structured output pipelines. Gathers requirements, designs step sequences with output schemas, generates valid YAML pipeline definitions, and validates them.","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002Fdocs\u002Fbuiltin-skills\u002Fpipeline-designer\u002FSKILL.md",[78],{"categoryNo":79,"categoryName":80,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000015","视觉设计",[82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,39,99,100],"pipeline designer","pipeline","designer","create","valid","design","requirements","file","yaml","structured","shell","message","what the","ask","question","read","write","pipelines","do",{"skillNo":102,"name":103,"nameZh":6,"description":104,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":105,"categories":106,"keywords":107,"publishedAt":43},"01M6CZGAY23A6HQS1G7JK7WQGM","idea","Log a planning idea to the backlog. Use when the user has a feature idea, enhancement suggestion, or brainstorm they want tracked.","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002Fdocs\u002Fbuiltin-skills\u002Fidea\u002FSKILL.md",[],[103,28,29,108,37,39,109,110],"log","brainstorm","planning",{"skillNo":112,"name":28,"nameZh":6,"description":113,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":114,"categories":115,"keywords":117,"publishedAt":43},"01M6CZG7S5TSVHB6S8YA3JMGFJ","Automate the creation of a new feature for clawq. Gathers requirements, discovers existing capabilities, decomposes into implementation layers, gates risky work behind approval, and delegates implementation. Prefers skills\u002Frigs\u002Fcron\u002Ftemplates over OCaml code changes.","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002Fdocs\u002Fbuiltin-skills\u002Ffeature\u002FSKILL.md",[116],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[28,118,119,120,121,122,123,124,125,126,127,128,129,130,88,131,132],"language","capabilities","code","clawq","templates","cron","automate","work","new feature","approval","implementation","ocaml","natural language","gates","creation",{"skillNo":134,"name":135,"nameZh":6,"description":136,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":137,"categories":138,"keywords":139,"publishedAt":43},"01M6CZG4STQ8R15F6HFT5BE98J","telegram-formatting","Telegram HTML and MarkdownV2 formatting reference for clawq message code","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002F.claude\u002Fskills\u002Ftelegram-formatting\u002FSKILL.md",[],[140,141,120,142,143,93,121],"telegram formatting","telegram","html","reference",{"skillNo":145,"name":146,"nameZh":6,"description":147,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":148,"categories":149,"keywords":150,"publishedAt":43},"01M6CZG2ZASD966C153XZY3BQ1","pag-review-via-team","review work using teams + review-and-fix","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002F.claude\u002Fskills\u002Fpag-review-via-team\u002FSKILL.md",[],[151,152,27,153,125,154],"pag review via team","team","teams","fix",{"skillNo":156,"name":157,"nameZh":6,"description":158,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":159,"categories":160,"keywords":161,"publishedAt":43},"01M6CZG02S14BW5NN1P2F2PN7Q","clawq-version-bump","Bump clawq version, commit, tag, push, and create a GitHub release","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002F.claude\u002Fskills\u002Fclawq-version-bump\u002FSKILL.md",[],[121,162,163,164,165,166,85,167,168,169,170],"clawq version bump","version bump","bump","version","commit","release","push","github release","github"]