[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6CZG7S5TSVHB6S8YA3JMGFJ":3,"skill-same:01M6CZG7S5TSVHB6S8YA3JMGFJ":68},{"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":39,"linkOnly":40,"markdownContent":41,"markdownBody":42,"frontMatter":43,"refName":46,"commitSha":47,"skillDirectory":48,"forksCount":13,"repositoryHtmlUrl":49,"repositoryHomepageUrl":50,"ownerHtmlUrl":51,"ownerBlogUrl":52,"ownerDescription":6,"updatedAt":53,"install":54,"resources":67},"01M6CZG7S5TSVHB6S8YA3JMGFJ","feature",null,"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.","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\u002Ffeature\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[5,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38],"implementation","clawq","creation","gates","requirements","natural language","language","ocaml","code","approval","new feature","work","automate","cron","templates","capabilities","2026-11-10 10:12:30",false,"---\nname: feature\ndescription: 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.\nargument-hint: [feature description in natural language]\n---\n\n# \u002Ffeature — Automated Feature Creation\n\nCreate a new clawq feature end-to-end: gather requirements, discover existing capabilities, decompose into implementation layers (preferring no-recompilation approaches), gate risky changes behind admin approval, and delegate implementation.\n\n## Mandatory Progress Protocol\n\n**You MUST emit a progress message at the start of every phase below before doing anything else in that phase.** Use `send_message(text=...)` — it delivers immediately to the active session. Failing to emit progress leaves the user staring at a silent terminal; treat this as a hard requirement, not a suggestion.\n\nThe six messages are exactly:\n\n1. `Add Feature, step 1\u002F6: Gathering requirements...`\n2. `Add Feature, step 2\u002F6: Discovering capabilities...`\n3. `Add Feature, step 3\u002F6: Decomposing feature...`\n4. `Add Feature, step 4\u002F6: Checking approvals...`\n5. `Add Feature, step 5\u002F6: Implementing...`\n6. `Add Feature, step 6\u002F6: Verifying and reporting...`\n\nAdditionally, emit a `send_message` whenever you actually load a skill, spawn a background task, or are about to wait on a long operation (>5 s). Do not emit a loading message when `use_skill` is a no-op because that no-argument skill is already present in current agent context. Examples:\n\n- `Loading skill: \u003Cname>`\n- `Spawning background task \u003Crunner> on branch \u003Cbranch>...`\n- `Waiting on background task \u003Cid> (this may take a few minutes)...`\n\nIf `send_message` is not registered (no active session), fall back to printing a plain assistant message that begins with the same text — but only after attempting send_message first.\n\n## Phase 1: Requirement Gathering\n\n**Step 1.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 1\u002F6: Gathering requirements...\")`.\n\nIf `$ARGUMENTS` provides a description, parse it for context and pre-fill answers. Confirm with the user.\n\nCollect via `ask_user_question` (or conversationally if unavailable):\n\n1. **Feature name** (text) — short identifier (e.g. \"daily-digest\")\n2. **Description** (text) — what the feature should do\n3. **Access method preference** (single_select: \"slash command\", \"scheduled task\", \"new tool\", \"CLI command\", \"let the skill decide\")\n4. **Acceptance criteria** (text) — how to verify it works\n5. **Priority** (single_select: \"low\", \"medium\", \"high\", \"critical\")\n\n## Phase 2: Capability Discovery\n\n**Step 2.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 2\u002F6: Discovering capabilities...\")`.\n\nRead existing capabilities:\n\n1. `file_read(\"docs\u002Fpublic\u002Fllms-full.txt\")` — full self-knowledge reference\n2. `skill_list()` — existing skills\n3. `shell_exec(\"clawq rig list\")` — existing rigs\n4. `shell_exec(\"clawq cron list\")` — existing cron jobs\n5. `shell_exec(\"clawq agents list\")` — existing agent templates\n6. `tool_search(\"\u003Ckeywords from feature description>\")` — deferred tool discovery\n\nAfter each `shell_exec` call above, emit `send_message(text=\"Checked: \u003Cthing>\")` so the user sees discovery progress.\n\nSummarize: \"Here's what clawq can already do related to your request: ...\"\n\n## Phase 3: Feature Decomposition\n\n**Step 3.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 3\u002F6: Decomposing feature...\")`.\n\nClassify each requirement into implementation layers (prefer lower layers):\n\n| Layer | What | Recompilation? | Approval |\n|-------|------|----------------|----------|\n| 0 | Already exists — use existing tool\u002Fskill\u002Fcommand | No | None |\n| 1 | SKILL.md — agent instructions + existing tools | No | None |\n| 2 | Rig — setup workflow + external tools | No | Confirm |\n| 3 | Skill + Cron — periodic execution | No | Confirm |\n| 4 | Agent Template — specialized persona | No | None |\n| 5 | MCP Server — new tool via external process | No | Confirm |\n| 6 | OCaml changes — core runtime modification | **Yes** | **Admin required** |\n\nPresent the decomposition table and ask for confirmation via `ask_user_question` (type: confirm):\n\"Here's how I'd implement this feature. Proceed?\"\n\n## Phase 4: Approval Gating\n\n**Step 4.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 4\u002F6: Checking approvals...\")`.\n\n**Layers 0-4** (no recompilation, low risk): Proceed after user confirmation.\n\n**Layer 5** (MCP server, medium risk): Confirm with user, explain what will be created.\n\n**Layer 6** (OCaml changes, high risk):\n\n1. Check if the user has admin privileges (from the session context).\n2. If **guest**: save the full feature plan for admin review:\n   - Write the plan JSON to a temp file\n   - Run: `shell_exec(\"clawq held-items save --name '\u003Cname>' --desc '\u003Cdescription>' --plan-file \u002Ftmp\u002Fheld_plan_\u003Ctimestamp>.json --layer 6 --requestor '\u003Csender_id>' --channel '\u003Cchannel>'\")`\n   - Respond: \"This feature requires OCaml code changes. Your plan has been saved as a held item for admin review. An admin can view and approve it via `\u002Fheld-items`.\"\n   - Stop here for guest users.\n3. If **admin**: present full impact summary (files to change, tests needed, estimated scope), then confirm via `ask_user_question`.\n4. Only after admin confirmation: proceed to Phase 5.\n\n## Phase 5: Implementation\n\n**Step 5.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 5\u002F6: Implementing...\")`.\n\nExecute in dependency order based on decomposition. Before each Layer block below, emit `send_message(text=\"Implementing layer \u003CN>: \u003Cshort description>\")`.\n\n### Layer 0 (already exists)\nShow usage instructions to the user. Done.\n\n### Layer 1 (SKILL.md)\n- Create the skill file: `file_write(\"~\u002F.clawq\u002Fskills\u002F\u003Cname>\u002FSKILL.md\", \"\u003Cskill content>\")`\n  - Include frontmatter with name, description, argument-hint\n  - Write clear step-by-step instructions the agent can follow\n- Test immediately: `use_skill(\"\u003Cname>\", arguments=\"test\")`\n- Report: \"Created skill `\u002Fskill \u003Cname>`. Use it in chat or it auto-triggers from matching keywords.\"\n\n### Layer 2 (Rig)\n- Write a rig definition markdown describing setup steps and configuration\n- Report: \"Created rig. Install via `clawq rig install \u003Cname>` or `\u002Frig install \u003Cname>`.\"\n\n### Layer 3 (Skill + Cron)\n- Create the skill first (Layer 1 flow)\n- Add cron job: `shell_exec(\"clawq cron add \u003Cname> \u003Csession> '\u003Cschedule>' '\u003Cmessage>' --ttl \u003Cduration>\")`\n- Report: \"Created skill + cron job. Runs on schedule: `\u003Cschedule>`. Manual trigger: `\u002Fskill \u003Cname>`.\"\n\n### Layer 4 (Agent Template)\n- Write template to `~\u002F.clawq\u002Fagents\u002F\u003Cname>.md`\n- Report: \"Created agent template `@\u003Cname>`. Use via `\u002Fdelegate @\u003Cname> \u003Cprompt>`.\"\n\n### Layer 5 (MCP Server)\n- Write a server script (Python or Node) implementing the MCP protocol\n- Register in `~\u002F.clawq\u002Fmcp_servers.json`\n- Report: \"Created MCP server. New tools will be available in agent context after restart.\"\n\n### Layer 6 (OCaml changes)\n1. Create backlog items: `shell_exec(\"bl idea '\u003Ctitle>' --body '\u003Cdetailed requirements and acceptance criteria>'\")`\n2. Ask user to choose execution strategy via `ask_user_question` (single_select):\n\n   **Small tasks** (single-component, straightforward):\n   - \"Quick (Kimi K2)\" — fast iteration\n   - \"Standard (Codex GPT-5.4)\" — reliable general-purpose\n   - \"Thorough (Claude Opus)\" — complex logic\n\n   **Large tasks** (multi-component, cross-cutting):\n   - \"Single coder\" — one agent, coherent single-pass\n   - \"Small team (orchestrator + coder + reviewer)\" — multi-file with review\n   - \"Full team (lead + N coders + reviewer)\" — cross-subsystem changes\n\n3. Before spawning, emit `send_message(text=\"Spawning background task \u003Crunner> on branch \u003Cbranch>...\")`.\n4. Spawn background task:\n   ```\n   background_task_enqueue(\n     runner: \"\u003Cselected, or local for a native subagent>\",\n     repo_path: \"\u003Crepo>\",\n     prompt: \"\u003Cconstructed prompt with requirements, AC, file targets>\",\n     branch: \"clawq-feature-\u003Cname>\",\n     use_worktree: true\n   )\n   ```\n5. Monitor: `background_task_logs(id)`, `background_task_wait(id)`, and `background_task_transcript(id, regex: \"...\")` for bounded conversation history. Local\u002Fnative subagents persist history under `__bg_task:\u003Cid>`.\n6. Steer: `background_task_send_message(id, message: \"...\")`\n7. On failure: `background_task_recover(id)`\n\n## Phase 6: Verification and Reporting\n\n**Step 6.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 6\u002F6: Verifying and reporting...\")`.\n\n1. **Skills**: invoke via `use_skill` to verify loading\n2. **Cron**: verify via `shell_exec(\"clawq cron list\")`\n3. **Rigs**: verify via `shell_exec(\"clawq rig list\")`\n4. **OCaml**: check `background_task_list` for completion, inspect `background_task_transcript` when context is needed, verify `make test` passed\n5. **Store**: `memory_store(key=\"feature:\u003Cname>\", value=\"\u003Csummary of what was created>\")`\n\nPresent final report:\n- What was created (with specific file paths)\n- How to access\u002Fuse it (specific commands the user can run)\n- What's pending (if OCaml implementation in progress)\n- Link to backlog item (if created)\n\n## Access Method Decision Matrix\n\n| Feature Type | Recommended Access |\n|---|---|\n| Workflow\u002Fautomation | `\u002Fskill \u003Cname> \u003Cargs>` in chat |\n| Periodic task | Automatic on schedule; manual via `\u002Fskill \u003Cname>` |\n| Setup\u002Fconfig | `\u002Frig install \u003Cname>` |\n| Specialized agent | `\u002Fdelegate @\u003Cname> \u003Cprompt>` |\n| External tool | Available as tool in agent context (via MCP) |\n| New CLI command | `clawq \u003Cname>` (after OCaml changes) |\n| New built-in tool | Available in agent tool list (after recompilation) |\n","# \u002Ffeature — Automated Feature Creation\n\nCreate a new clawq feature end-to-end: gather requirements, discover existing capabilities, decompose into implementation layers (preferring no-recompilation approaches), gate risky changes behind admin approval, and delegate implementation.\n\n## Mandatory Progress Protocol\n\n**You MUST emit a progress message at the start of every phase below before doing anything else in that phase.** Use `send_message(text=...)` — it delivers immediately to the active session. Failing to emit progress leaves the user staring at a silent terminal; treat this as a hard requirement, not a suggestion.\n\nThe six messages are exactly:\n\n1. `Add Feature, step 1\u002F6: Gathering requirements...`\n2. `Add Feature, step 2\u002F6: Discovering capabilities...`\n3. `Add Feature, step 3\u002F6: Decomposing feature...`\n4. `Add Feature, step 4\u002F6: Checking approvals...`\n5. `Add Feature, step 5\u002F6: Implementing...`\n6. `Add Feature, step 6\u002F6: Verifying and reporting...`\n\nAdditionally, emit a `send_message` whenever you actually load a skill, spawn a background task, or are about to wait on a long operation (>5 s). Do not emit a loading message when `use_skill` is a no-op because that no-argument skill is already present in current agent context. Examples:\n\n- `Loading skill: \u003Cname>`\n- `Spawning background task \u003Crunner> on branch \u003Cbranch>...`\n- `Waiting on background task \u003Cid> (this may take a few minutes)...`\n\nIf `send_message` is not registered (no active session), fall back to printing a plain assistant message that begins with the same text — but only after attempting send_message first.\n\n## Phase 1: Requirement Gathering\n\n**Step 1.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 1\u002F6: Gathering requirements...\")`.\n\nIf `$ARGUMENTS` provides a description, parse it for context and pre-fill answers. Confirm with the user.\n\nCollect via `ask_user_question` (or conversationally if unavailable):\n\n1. **Feature name** (text) — short identifier (e.g. \"daily-digest\")\n2. **Description** (text) — what the feature should do\n3. **Access method preference** (single_select: \"slash command\", \"scheduled task\", \"new tool\", \"CLI command\", \"let the skill decide\")\n4. **Acceptance criteria** (text) — how to verify it works\n5. **Priority** (single_select: \"low\", \"medium\", \"high\", \"critical\")\n\n## Phase 2: Capability Discovery\n\n**Step 2.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 2\u002F6: Discovering capabilities...\")`.\n\nRead existing capabilities:\n\n1. `file_read(\"docs\u002Fpublic\u002Fllms-full.txt\")` — full self-knowledge reference\n2. `skill_list()` — existing skills\n3. `shell_exec(\"clawq rig list\")` — existing rigs\n4. `shell_exec(\"clawq cron list\")` — existing cron jobs\n5. `shell_exec(\"clawq agents list\")` — existing agent templates\n6. `tool_search(\"\u003Ckeywords from feature description>\")` — deferred tool discovery\n\nAfter each `shell_exec` call above, emit `send_message(text=\"Checked: \u003Cthing>\")` so the user sees discovery progress.\n\nSummarize: \"Here's what clawq can already do related to your request: ...\"\n\n## Phase 3: Feature Decomposition\n\n**Step 3.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 3\u002F6: Decomposing feature...\")`.\n\nClassify each requirement into implementation layers (prefer lower layers):\n\n| Layer | What | Recompilation? | Approval |\n|-------|------|----------------|----------|\n| 0 | Already exists — use existing tool\u002Fskill\u002Fcommand | No | None |\n| 1 | SKILL.md — agent instructions + existing tools | No | None |\n| 2 | Rig — setup workflow + external tools | No | Confirm |\n| 3 | Skill + Cron — periodic execution | No | Confirm |\n| 4 | Agent Template — specialized persona | No | None |\n| 5 | MCP Server — new tool via external process | No | Confirm |\n| 6 | OCaml changes — core runtime modification | **Yes** | **Admin required** |\n\nPresent the decomposition table and ask for confirmation via `ask_user_question` (type: confirm):\n\"Here's how I'd implement this feature. Proceed?\"\n\n## Phase 4: Approval Gating\n\n**Step 4.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 4\u002F6: Checking approvals...\")`.\n\n**Layers 0-4** (no recompilation, low risk): Proceed after user confirmation.\n\n**Layer 5** (MCP server, medium risk): Confirm with user, explain what will be created.\n\n**Layer 6** (OCaml changes, high risk):\n\n1. Check if the user has admin privileges (from the session context).\n2. If **guest**: save the full feature plan for admin review:\n   - Write the plan JSON to a temp file\n   - Run: `shell_exec(\"clawq held-items save --name '\u003Cname>' --desc '\u003Cdescription>' --plan-file \u002Ftmp\u002Fheld_plan_\u003Ctimestamp>.json --layer 6 --requestor '\u003Csender_id>' --channel '\u003Cchannel>'\")`\n   - Respond: \"This feature requires OCaml code changes. Your plan has been saved as a held item for admin review. An admin can view and approve it via `\u002Fheld-items`.\"\n   - Stop here for guest users.\n3. If **admin**: present full impact summary (files to change, tests needed, estimated scope), then confirm via `ask_user_question`.\n4. Only after admin confirmation: proceed to Phase 5.\n\n## Phase 5: Implementation\n\n**Step 5.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 5\u002F6: Implementing...\")`.\n\nExecute in dependency order based on decomposition. Before each Layer block below, emit `send_message(text=\"Implementing layer \u003CN>: \u003Cshort description>\")`.\n\n### Layer 0 (already exists)\nShow usage instructions to the user. Done.\n\n### Layer 1 (SKILL.md)\n- Create the skill file: `file_write(\"~\u002F.clawq\u002Fskills\u002F\u003Cname>\u002FSKILL.md\", \"\u003Cskill content>\")`\n  - Include frontmatter with name, description, argument-hint\n  - Write clear step-by-step instructions the agent can follow\n- Test immediately: `use_skill(\"\u003Cname>\", arguments=\"test\")`\n- Report: \"Created skill `\u002Fskill \u003Cname>`. Use it in chat or it auto-triggers from matching keywords.\"\n\n### Layer 2 (Rig)\n- Write a rig definition markdown describing setup steps and configuration\n- Report: \"Created rig. Install via `clawq rig install \u003Cname>` or `\u002Frig install \u003Cname>`.\"\n\n### Layer 3 (Skill + Cron)\n- Create the skill first (Layer 1 flow)\n- Add cron job: `shell_exec(\"clawq cron add \u003Cname> \u003Csession> '\u003Cschedule>' '\u003Cmessage>' --ttl \u003Cduration>\")`\n- Report: \"Created skill + cron job. Runs on schedule: `\u003Cschedule>`. Manual trigger: `\u002Fskill \u003Cname>`.\"\n\n### Layer 4 (Agent Template)\n- Write template to `~\u002F.clawq\u002Fagents\u002F\u003Cname>.md`\n- Report: \"Created agent template `@\u003Cname>`. Use via `\u002Fdelegate @\u003Cname> \u003Cprompt>`.\"\n\n### Layer 5 (MCP Server)\n- Write a server script (Python or Node) implementing the MCP protocol\n- Register in `~\u002F.clawq\u002Fmcp_servers.json`\n- Report: \"Created MCP server. New tools will be available in agent context after restart.\"\n\n### Layer 6 (OCaml changes)\n1. Create backlog items: `shell_exec(\"bl idea '\u003Ctitle>' --body '\u003Cdetailed requirements and acceptance criteria>'\")`\n2. Ask user to choose execution strategy via `ask_user_question` (single_select):\n\n   **Small tasks** (single-component, straightforward):\n   - \"Quick (Kimi K2)\" — fast iteration\n   - \"Standard (Codex GPT-5.4)\" — reliable general-purpose\n   - \"Thorough (Claude Opus)\" — complex logic\n\n   **Large tasks** (multi-component, cross-cutting):\n   - \"Single coder\" — one agent, coherent single-pass\n   - \"Small team (orchestrator + coder + reviewer)\" — multi-file with review\n   - \"Full team (lead + N coders + reviewer)\" — cross-subsystem changes\n\n3. Before spawning, emit `send_message(text=\"Spawning background task \u003Crunner> on branch \u003Cbranch>...\")`.\n4. Spawn background task:\n   ```\n   background_task_enqueue(\n     runner: \"\u003Cselected, or local for a native subagent>\",\n     repo_path: \"\u003Crepo>\",\n     prompt: \"\u003Cconstructed prompt with requirements, AC, file targets>\",\n     branch: \"clawq-feature-\u003Cname>\",\n     use_worktree: true\n   )\n   ```\n5. Monitor: `background_task_logs(id)`, `background_task_wait(id)`, and `background_task_transcript(id, regex: \"...\")` for bounded conversation history. Local\u002Fnative subagents persist history under `__bg_task:\u003Cid>`.\n6. Steer: `background_task_send_message(id, message: \"...\")`\n7. On failure: `background_task_recover(id)`\n\n## Phase 6: Verification and Reporting\n\n**Step 6.0 (MANDATORY):** call `send_message(text=\"Add Feature, step 6\u002F6: Verifying and reporting...\")`.\n\n1. **Skills**: invoke via `use_skill` to verify loading\n2. **Cron**: verify via `shell_exec(\"clawq cron list\")`\n3. **Rigs**: verify via `shell_exec(\"clawq rig list\")`\n4. **OCaml**: check `background_task_list` for completion, inspect `background_task_transcript` when context is needed, verify `make test` passed\n5. **Store**: `memory_store(key=\"feature:\u003Cname>\", value=\"\u003Csummary of what was created>\")`\n\nPresent final report:\n- What was created (with specific file paths)\n- How to access\u002Fuse it (specific commands the user can run)\n- What's pending (if OCaml implementation in progress)\n- Link to backlog item (if created)\n\n## Access Method Decision Matrix\n\n| Feature Type | Recommended Access |\n|---|---|\n| Workflow\u002Fautomation | `\u002Fskill \u003Cname> \u003Cargs>` in chat |\n| Periodic task | Automatic on schedule; manual via `\u002Fskill \u003Cname>` |\n| Setup\u002Fconfig | `\u002Frig install \u003Cname>` |\n| Specialized agent | `\u002Fdelegate @\u003Cname> \u003Cprompt>` |\n| External tool | Available as tool in agent context (via MCP) |\n| New CLI command | `clawq \u003Cname>` (after OCaml changes) |\n| New built-in tool | Available in agent tool list (after recompilation) |\n",{"name":5,"description":7,"argument-hint":44},[45],"feature description in natural language","master","7c87c0cf8d869c392efbe4b2f0065c85657bf969","docs\u002Fbuiltin-skills\u002Ffeature","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":55,"prompt":65,"zipAvailable":66},[56,59,62],{"tool":57,"command":58},"npx","npx skills add clankercode\u002Fclawq --skill feature",{"tool":60,"command":61},"bunx","bunx skills add clankercode\u002Fclawq --skill feature",{"tool":63,"command":64},"pnpm","pnpm dlx skills add clankercode\u002Fclawq --skill feature","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002Fdocs\u002Fbuiltin-skills\u002Ffeature\u002FSKILL.md",true,[],[69,96,124,134,145,156],{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":76,"publishedAt":39},"01M6CZGG2GJZ7PGVTZSG8NDB2R","run-task","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.","https:\u002F\u002Fgithub.com\u002Fclankercode\u002Fclawq\u002Fblob\u002Fmaster\u002Fdocs\u002Fbuiltin-skills\u002Frun-task\u002FSKILL.md",[75],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,5],"run task","run","task","worktree","review","startup","subagent","extra","prompt","user","execute","use","bug","cycle","milestone","claim","done","review cycle","backlog",{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":105,"publishedAt":39},"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",[102],{"categoryNo":103,"categoryName":104,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000015","视觉设计",[106,107,108,109,110,86,111,27,112,113,114,115,116,117,118,119,120,121,122,123],"pipeline designer","pipeline","designer","write","yaml","pipelines","structured","message","what the","do","read","design","valid","create","file","shell","ask","question",{"skillNo":125,"name":126,"nameZh":6,"description":127,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":128,"categories":129,"keywords":130,"publishedAt":39},"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",[],[126,131,132,86,88,133,95,5],"brainstorm","planning","log",{"skillNo":135,"name":136,"nameZh":6,"description":137,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":138,"categories":139,"keywords":140,"publishedAt":39},"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",[],[141,142,31,143,144,113,24],"telegram formatting","telegram","html","reference",{"skillNo":146,"name":147,"nameZh":6,"description":148,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":149,"categories":150,"keywords":151,"publishedAt":39},"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",[],[152,153,81,154,34,155],"pag review via team","team","teams","fix",{"skillNo":157,"name":158,"nameZh":6,"description":159,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":160,"categories":161,"keywords":162,"publishedAt":39},"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",[],[163,24,164,165,166,167,119,168,169,170,171],"clawq version bump","version bump","bump","version","commit","release","push","github release","github"]