[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6KSTC0AJFNEZPKAS1PFC61N":3,"skill-same:01M6KSTC0AJFNEZPKAS1PFC61N":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":17,"publishedAt":38,"linkOnly":39,"markdownContent":40,"markdownBody":41,"frontMatter":42,"refName":48,"commitSha":49,"skillDirectory":50,"forksCount":51,"repositoryHtmlUrl":52,"repositoryHomepageUrl":6,"ownerHtmlUrl":53,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":54,"install":55,"resources":67},"01M6KSTC0AJFNEZPKAS1PFC61N","speckit-clarify",null,"Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.","ghbvf\u002Fgocell","ghbvf","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F104540935?v=4","01M6KSS2CVFY1XK013726EE7ZP","01M6KSS2GJ30TNAVPNFATW578Q",4,"MIT","https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.claude\u002Fskills\u002Fspeckit-clarify\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37],"speckit","speckit clarify","clarify","structure","true","kit","questions","with","spec-kit","feature","commands","project","directory","back","up","templates","specify","project structure","github","spec","2026-11-10 10:12:30",false,"---\nname: speckit-clarify\ndescription: Identify underspecified areas in the current feature spec by asking up\n  to 5 highly targeted clarification questions and encoding answers back into the\n  spec.\ncompatibility: Requires spec-kit project structure with .specify\u002F directory\nmetadata:\n  author: github-spec-kit\n  source: templates\u002Fcommands\u002Fclarify.md\ndisable-model-invocation: true\n---\n\n## User Input\n\n```text\n$ARGUMENTS\n```\n\nYou **MUST** consider the user input before proceeding (if not empty).\n\n## Outline\n\nGoal: Detect and reduce ambiguity or missing decision points in the active feature specification and record the clarifications directly in the spec file.\n\nNote: This clarification workflow is expected to run (and be completed) BEFORE invoking `\u002Fspeckit.plan`. If the user explicitly states they are skipping clarification (e.g., exploratory spike), you may proceed, but must warn that downstream rework risk increases.\n\nExecution steps:\n\n1. Run `.specify\u002Fscripts\u002Fbash\u002Fcheck-prerequisites.sh --json --paths-only` from repo root **once** (combined `--json --paths-only` mode \u002F `-Json -PathsOnly`). Parse minimal JSON payload fields:\n   - `FEATURE_DIR`\n   - `FEATURE_SPEC`\n   - (Optionally capture `IMPL_PLAN`, `TASKS` for future chained flows.)\n   - If JSON parsing fails, abort and instruct user to re-run `\u002Fspeckit.specify` or verify feature branch environment.\n   - For single quotes in args like \"I'm Groot\", use escape syntax: e.g 'I'\\''m Groot' (or double-quote if possible: \"I'm Groot\").\n\n2. Load the current spec file. Perform a structured ambiguity & coverage scan using this taxonomy. For each category, mark status: Clear \u002F Partial \u002F Missing. Produce an internal coverage map used for prioritization (do not output raw map unless no questions will be asked).\n\n   Functional Scope & Behavior:\n   - Core user goals & success criteria\n   - Explicit out-of-scope declarations\n   - User roles \u002F personas differentiation\n\n   Domain & Data Model:\n   - Entities, attributes, relationships\n   - Identity & uniqueness rules\n   - Lifecycle\u002Fstate transitions\n   - Data volume \u002F scale assumptions\n\n   Interaction & UX Flow:\n   - Critical user journeys \u002F sequences\n   - Error\u002Fempty\u002Floading states\n   - Accessibility or localization notes\n\n   Non-Functional Quality Attributes:\n   - Performance (latency, throughput targets)\n   - Scalability (horizontal\u002Fvertical, limits)\n   - Reliability & availability (uptime, recovery expectations)\n   - Observability (logging, metrics, tracing signals)\n   - Security & privacy (authN\u002FZ, data protection, threat assumptions)\n   - Compliance \u002F regulatory constraints (if any)\n\n   Integration & External Dependencies:\n   - External services\u002FAPIs and failure modes\n   - Data import\u002Fexport formats\n   - Protocol\u002Fversioning assumptions\n\n   Edge Cases & Failure Handling:\n   - Negative scenarios\n   - Rate limiting \u002F throttling\n   - Conflict resolution (e.g., concurrent edits)\n\n   Constraints & Tradeoffs:\n   - Technical constraints (language, storage, hosting)\n   - Explicit tradeoffs or rejected alternatives\n\n   Terminology & Consistency:\n   - Canonical glossary terms\n   - Avoided synonyms \u002F deprecated terms\n\n   Completion Signals:\n   - Acceptance criteria testability\n   - Measurable Definition of Done style indicators\n\n   Misc \u002F Placeholders:\n   - TODO markers \u002F unresolved decisions\n   - Ambiguous adjectives (\"robust\", \"intuitive\") lacking quantification\n\n   For each category with Partial or Missing status, add a candidate question opportunity unless:\n   - Clarification would not materially change implementation or validation strategy\n   - Information is better deferred to planning phase (note internally)\n\n3. Generate (internally) a prioritized queue of candidate clarification questions (maximum 5). Do NOT output them all at once. Apply these constraints:\n    - Maximum of 5 total questions across the whole session.\n    - Each question must be answerable with EITHER:\n       - A short multiple‑choice selection (2–5 distinct, mutually exclusive options), OR\n       - A one-word \u002F short‑phrase answer (explicitly constrain: \"Answer in \u003C=5 words\").\n    - Only include questions whose answers materially impact architecture, data modeling, task decomposition, test design, UX behavior, operational readiness, or compliance validation.\n    - Ensure category coverage balance: attempt to cover the highest impact unresolved categories first; avoid asking two low-impact questions when a single high-impact area (e.g., security posture) is unresolved.\n    - Exclude questions already answered, trivial stylistic preferences, or plan-level execution details (unless blocking correctness).\n    - Favor clarifications that reduce downstream rework risk or prevent misaligned acceptance tests.\n    - If more than 5 categories remain unresolved, select the top 5 by (Impact * Uncertainty) heuristic.\n\n4. Sequential questioning loop (interactive):\n    - Present EXACTLY ONE question at a time.\n    - For multiple‑choice questions:\n       - **Analyze all options** and determine the **most suitable option** based on:\n          - Best practices for the project type\n          - Common patterns in similar implementations\n          - Risk reduction (security, performance, maintainability)\n          - Alignment with any explicit project goals or constraints visible in the spec\n       - Present your **recommended option prominently** at the top with clear reasoning (1-2 sentences explaining why this is the best choice).\n       - Format as: `**Recommended:** Option [X] - \u003Creasoning>`\n       - Then render all options as a Markdown table:\n\n       | Option | Description |\n       |--------|-------------|\n       | A | \u003COption A description> |\n       | B | \u003COption B description> |\n       | C | \u003COption C description> (add D\u002FE as needed up to 5) |\n       | Short | Provide a different short answer (\u003C=5 words) (Include only if free-form alternative is appropriate) |\n\n       - After the table, add: `You can reply with the option letter (e.g., \"A\"), accept the recommendation by saying \"yes\" or \"recommended\", or provide your own short answer.`\n    - For short‑answer style (no meaningful discrete options):\n       - Provide your **suggested answer** based on best practices and context.\n       - Format as: `**Suggested:** \u003Cyour proposed answer> - \u003Cbrief reasoning>`\n       - Then output: `Format: Short answer (\u003C=5 words). You can accept the suggestion by saying \"yes\" or \"suggested\", or provide your own answer.`\n    - After the user answers:\n       - If the user replies with \"yes\", \"recommended\", or \"suggested\", use your previously stated recommendation\u002Fsuggestion as the answer.\n       - Otherwise, validate the answer maps to one option or fits the \u003C=5 word constraint.\n       - If ambiguous, ask for a quick disambiguation (count still belongs to same question; do not advance).\n       - Once satisfactory, record it in working memory (do not yet write to disk) and move to the next queued question.\n    - Stop asking further questions when:\n       - All critical ambiguities resolved early (remaining queued items become unnecessary), OR\n       - User signals completion (\"done\", \"good\", \"no more\"), OR\n       - You reach 5 asked questions.\n    - Never reveal future queued questions in advance.\n    - If no valid questions exist at start, immediately report no critical ambiguities.\n\n5. Integration after EACH accepted answer (incremental update approach):\n    - Maintain in-memory representation of the spec (loaded once at start) plus the raw file contents.\n    - For the first integrated answer in this session:\n       - Ensure a `## Clarifications` section exists (create it just after the highest-level contextual\u002Foverview section per the spec template if missing).\n       - Under it, create (if not present) a `### Session YYYY-MM-DD` subheading for today.\n    - Append a bullet line immediately after acceptance: `- Q: \u003Cquestion> → A: \u003Cfinal answer>`.\n    - Then immediately apply the clarification to the most appropriate section(s):\n       - Functional ambiguity → Update or add a bullet in Functional Requirements.\n       - User interaction \u002F actor distinction → Update User Stories or Actors subsection (if present) with clarified role, constraint, or scenario.\n       - Data shape \u002F entities → Update Data Model (add fields, types, relationships) preserving ordering; note added constraints succinctly.\n       - Non-functional constraint → Add\u002Fmodify measurable criteria in Success Criteria > Measurable Outcomes (convert vague adjective to metric or explicit target).\n       - Edge case \u002F negative flow → Add a new bullet under Edge Cases \u002F Error Handling (or create such subsection if template provides placeholder for it).\n       - Terminology conflict → Normalize term across spec; retain original only if necessary by adding `(formerly referred to as \"X\")` once.\n    - If the clarification invalidates an earlier ambiguous statement, replace that statement instead of duplicating; leave no obsolete contradictory text.\n    - Save the spec file AFTER each integration to minimize risk of context loss (atomic overwrite).\n    - Preserve formatting: do not reorder unrelated sections; keep heading hierarchy intact.\n    - Keep each inserted clarification minimal and testable (avoid narrative drift).\n\n6. Validation (performed after EACH write plus final pass):\n   - Clarifications session contains exactly one bullet per accepted answer (no duplicates).\n   - Total asked (accepted) questions ≤ 5.\n   - Updated sections contain no lingering vague placeholders the new answer was meant to resolve.\n   - No contradictory earlier statement remains (scan for now-invalid alternative choices removed).\n   - Markdown structure valid; only allowed new headings: `## Clarifications`, `### Session YYYY-MM-DD`.\n   - Terminology consistency: same canonical term used across all updated sections.\n\n7. Write the updated spec back to `FEATURE_SPEC`.\n\n8. Report completion (after questioning loop ends or early termination):\n   - Number of questions asked & answered.\n   - Path to updated spec.\n   - Sections touched (list names).\n   - Coverage summary table listing each taxonomy category with Status: Resolved (was Partial\u002FMissing and addressed), Deferred (exceeds question quota or better suited for planning), Clear (already sufficient), Outstanding (still Partial\u002FMissing but low impact).\n   - If any Outstanding or Deferred remain, recommend whether to proceed to `\u002Fspeckit.plan` or run `\u002Fspeckit.clarify` again later post-plan.\n   - Suggested next command.\n\nBehavior rules:\n\n- If no meaningful ambiguities found (or all potential questions would be low-impact), respond: \"No critical ambiguities detected worth formal clarification.\" and suggest proceeding.\n- If spec file missing, instruct user to run `\u002Fspeckit.specify` first (do not create a new spec here).\n- Never exceed 5 total asked questions (clarification retries for a single question do not count as new questions).\n- Avoid speculative tech stack questions unless the absence blocks functional clarity.\n- Respect user early termination signals (\"stop\", \"done\", \"proceed\").\n- If no questions asked due to full coverage, output a compact coverage summary (all categories Clear) then suggest advancing.\n- If quota reached with unresolved high-impact categories remaining, explicitly flag them under Deferred with rationale.\n\nContext for prioritization: $ARGUMENTS\n","## User Input\n\n```text\n$ARGUMENTS\n```\n\nYou **MUST** consider the user input before proceeding (if not empty).\n\n## Outline\n\nGoal: Detect and reduce ambiguity or missing decision points in the active feature specification and record the clarifications directly in the spec file.\n\nNote: This clarification workflow is expected to run (and be completed) BEFORE invoking `\u002Fspeckit.plan`. If the user explicitly states they are skipping clarification (e.g., exploratory spike), you may proceed, but must warn that downstream rework risk increases.\n\nExecution steps:\n\n1. Run `.specify\u002Fscripts\u002Fbash\u002Fcheck-prerequisites.sh --json --paths-only` from repo root **once** (combined `--json --paths-only` mode \u002F `-Json -PathsOnly`). Parse minimal JSON payload fields:\n   - `FEATURE_DIR`\n   - `FEATURE_SPEC`\n   - (Optionally capture `IMPL_PLAN`, `TASKS` for future chained flows.)\n   - If JSON parsing fails, abort and instruct user to re-run `\u002Fspeckit.specify` or verify feature branch environment.\n   - For single quotes in args like \"I'm Groot\", use escape syntax: e.g 'I'\\''m Groot' (or double-quote if possible: \"I'm Groot\").\n\n2. Load the current spec file. Perform a structured ambiguity & coverage scan using this taxonomy. For each category, mark status: Clear \u002F Partial \u002F Missing. Produce an internal coverage map used for prioritization (do not output raw map unless no questions will be asked).\n\n   Functional Scope & Behavior:\n   - Core user goals & success criteria\n   - Explicit out-of-scope declarations\n   - User roles \u002F personas differentiation\n\n   Domain & Data Model:\n   - Entities, attributes, relationships\n   - Identity & uniqueness rules\n   - Lifecycle\u002Fstate transitions\n   - Data volume \u002F scale assumptions\n\n   Interaction & UX Flow:\n   - Critical user journeys \u002F sequences\n   - Error\u002Fempty\u002Floading states\n   - Accessibility or localization notes\n\n   Non-Functional Quality Attributes:\n   - Performance (latency, throughput targets)\n   - Scalability (horizontal\u002Fvertical, limits)\n   - Reliability & availability (uptime, recovery expectations)\n   - Observability (logging, metrics, tracing signals)\n   - Security & privacy (authN\u002FZ, data protection, threat assumptions)\n   - Compliance \u002F regulatory constraints (if any)\n\n   Integration & External Dependencies:\n   - External services\u002FAPIs and failure modes\n   - Data import\u002Fexport formats\n   - Protocol\u002Fversioning assumptions\n\n   Edge Cases & Failure Handling:\n   - Negative scenarios\n   - Rate limiting \u002F throttling\n   - Conflict resolution (e.g., concurrent edits)\n\n   Constraints & Tradeoffs:\n   - Technical constraints (language, storage, hosting)\n   - Explicit tradeoffs or rejected alternatives\n\n   Terminology & Consistency:\n   - Canonical glossary terms\n   - Avoided synonyms \u002F deprecated terms\n\n   Completion Signals:\n   - Acceptance criteria testability\n   - Measurable Definition of Done style indicators\n\n   Misc \u002F Placeholders:\n   - TODO markers \u002F unresolved decisions\n   - Ambiguous adjectives (\"robust\", \"intuitive\") lacking quantification\n\n   For each category with Partial or Missing status, add a candidate question opportunity unless:\n   - Clarification would not materially change implementation or validation strategy\n   - Information is better deferred to planning phase (note internally)\n\n3. Generate (internally) a prioritized queue of candidate clarification questions (maximum 5). Do NOT output them all at once. Apply these constraints:\n    - Maximum of 5 total questions across the whole session.\n    - Each question must be answerable with EITHER:\n       - A short multiple‑choice selection (2–5 distinct, mutually exclusive options), OR\n       - A one-word \u002F short‑phrase answer (explicitly constrain: \"Answer in \u003C=5 words\").\n    - Only include questions whose answers materially impact architecture, data modeling, task decomposition, test design, UX behavior, operational readiness, or compliance validation.\n    - Ensure category coverage balance: attempt to cover the highest impact unresolved categories first; avoid asking two low-impact questions when a single high-impact area (e.g., security posture) is unresolved.\n    - Exclude questions already answered, trivial stylistic preferences, or plan-level execution details (unless blocking correctness).\n    - Favor clarifications that reduce downstream rework risk or prevent misaligned acceptance tests.\n    - If more than 5 categories remain unresolved, select the top 5 by (Impact * Uncertainty) heuristic.\n\n4. Sequential questioning loop (interactive):\n    - Present EXACTLY ONE question at a time.\n    - For multiple‑choice questions:\n       - **Analyze all options** and determine the **most suitable option** based on:\n          - Best practices for the project type\n          - Common patterns in similar implementations\n          - Risk reduction (security, performance, maintainability)\n          - Alignment with any explicit project goals or constraints visible in the spec\n       - Present your **recommended option prominently** at the top with clear reasoning (1-2 sentences explaining why this is the best choice).\n       - Format as: `**Recommended:** Option [X] - \u003Creasoning>`\n       - Then render all options as a Markdown table:\n\n       | Option | Description |\n       |--------|-------------|\n       | A | \u003COption A description> |\n       | B | \u003COption B description> |\n       | C | \u003COption C description> (add D\u002FE as needed up to 5) |\n       | Short | Provide a different short answer (\u003C=5 words) (Include only if free-form alternative is appropriate) |\n\n       - After the table, add: `You can reply with the option letter (e.g., \"A\"), accept the recommendation by saying \"yes\" or \"recommended\", or provide your own short answer.`\n    - For short‑answer style (no meaningful discrete options):\n       - Provide your **suggested answer** based on best practices and context.\n       - Format as: `**Suggested:** \u003Cyour proposed answer> - \u003Cbrief reasoning>`\n       - Then output: `Format: Short answer (\u003C=5 words). You can accept the suggestion by saying \"yes\" or \"suggested\", or provide your own answer.`\n    - After the user answers:\n       - If the user replies with \"yes\", \"recommended\", or \"suggested\", use your previously stated recommendation\u002Fsuggestion as the answer.\n       - Otherwise, validate the answer maps to one option or fits the \u003C=5 word constraint.\n       - If ambiguous, ask for a quick disambiguation (count still belongs to same question; do not advance).\n       - Once satisfactory, record it in working memory (do not yet write to disk) and move to the next queued question.\n    - Stop asking further questions when:\n       - All critical ambiguities resolved early (remaining queued items become unnecessary), OR\n       - User signals completion (\"done\", \"good\", \"no more\"), OR\n       - You reach 5 asked questions.\n    - Never reveal future queued questions in advance.\n    - If no valid questions exist at start, immediately report no critical ambiguities.\n\n5. Integration after EACH accepted answer (incremental update approach):\n    - Maintain in-memory representation of the spec (loaded once at start) plus the raw file contents.\n    - For the first integrated answer in this session:\n       - Ensure a `## Clarifications` section exists (create it just after the highest-level contextual\u002Foverview section per the spec template if missing).\n       - Under it, create (if not present) a `### Session YYYY-MM-DD` subheading for today.\n    - Append a bullet line immediately after acceptance: `- Q: \u003Cquestion> → A: \u003Cfinal answer>`.\n    - Then immediately apply the clarification to the most appropriate section(s):\n       - Functional ambiguity → Update or add a bullet in Functional Requirements.\n       - User interaction \u002F actor distinction → Update User Stories or Actors subsection (if present) with clarified role, constraint, or scenario.\n       - Data shape \u002F entities → Update Data Model (add fields, types, relationships) preserving ordering; note added constraints succinctly.\n       - Non-functional constraint → Add\u002Fmodify measurable criteria in Success Criteria > Measurable Outcomes (convert vague adjective to metric or explicit target).\n       - Edge case \u002F negative flow → Add a new bullet under Edge Cases \u002F Error Handling (or create such subsection if template provides placeholder for it).\n       - Terminology conflict → Normalize term across spec; retain original only if necessary by adding `(formerly referred to as \"X\")` once.\n    - If the clarification invalidates an earlier ambiguous statement, replace that statement instead of duplicating; leave no obsolete contradictory text.\n    - Save the spec file AFTER each integration to minimize risk of context loss (atomic overwrite).\n    - Preserve formatting: do not reorder unrelated sections; keep heading hierarchy intact.\n    - Keep each inserted clarification minimal and testable (avoid narrative drift).\n\n6. Validation (performed after EACH write plus final pass):\n   - Clarifications session contains exactly one bullet per accepted answer (no duplicates).\n   - Total asked (accepted) questions ≤ 5.\n   - Updated sections contain no lingering vague placeholders the new answer was meant to resolve.\n   - No contradictory earlier statement remains (scan for now-invalid alternative choices removed).\n   - Markdown structure valid; only allowed new headings: `## Clarifications`, `### Session YYYY-MM-DD`.\n   - Terminology consistency: same canonical term used across all updated sections.\n\n7. Write the updated spec back to `FEATURE_SPEC`.\n\n8. Report completion (after questioning loop ends or early termination):\n   - Number of questions asked & answered.\n   - Path to updated spec.\n   - Sections touched (list names).\n   - Coverage summary table listing each taxonomy category with Status: Resolved (was Partial\u002FMissing and addressed), Deferred (exceeds question quota or better suited for planning), Clear (already sufficient), Outstanding (still Partial\u002FMissing but low impact).\n   - If any Outstanding or Deferred remain, recommend whether to proceed to `\u002Fspeckit.plan` or run `\u002Fspeckit.clarify` again later post-plan.\n   - Suggested next command.\n\nBehavior rules:\n\n- If no meaningful ambiguities found (or all potential questions would be low-impact), respond: \"No critical ambiguities detected worth formal clarification.\" and suggest proceeding.\n- If spec file missing, instruct user to run `\u002Fspeckit.specify` first (do not create a new spec here).\n- Never exceed 5 total asked questions (clarification retries for a single question do not count as new questions).\n- Avoid speculative tech stack questions unless the absence blocks functional clarity.\n- Respect user early termination signals (\"stop\", \"done\", \"proceed\").\n- If no questions asked due to full coverage, output a compact coverage summary (all categories Clear) then suggest advancing.\n- If quota reached with unresolved high-impact categories remaining, explicitly flag them under Deferred with rationale.\n\nContext for prioritization: $ARGUMENTS\n",{"name":5,"metadata":43,"description":7,"compatibility":46,"disable-model-invocation":47},{"author":44,"source":45},"github-spec-kit","templates\u002Fcommands\u002Fclarify.md","Requires spec-kit project structure with .specify\u002F directory",true,"develop","043add59bbe840e4a90a4253feec21f074b590dd",".claude\u002Fskills\u002Fspeckit-clarify",0,"https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell","https:\u002F\u002Fgithub.com\u002Fghbvf","2026-06-20 14:38:06",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":47},[57,60,63],{"tool":58,"command":59},"npx","npx skills add ghbvf\u002Fgocell --skill speckit-clarify",{"tool":61,"command":62},"bunx","bunx skills add ghbvf\u002Fgocell --skill speckit-clarify",{"tool":64,"command":65},"pnpm","pnpm dlx skills add ghbvf\u002Fgocell --skill speckit-clarify","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.claude\u002Fskills\u002Fspeckit-clarify\u002FSKILL.md",[],[69,90,105,124,134,141],{"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":75,"publishedAt":38},"01M6KSVEVMZN9HK0ZRVY3MEJEJ","ship","全流程实施：探索→计划→worktree→TDD→实施→PR→review→\u002Ffix Cx1\u002FCx2→人工确认。L1(跳过探索,1 reviewer)\u002FL2(单agent探索,1 reviewer)\u002FL3(默认,三agent探索,按diff行数1\u002F2\u002F3\u002F6 reviewer自动)","https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.codex\u002Fskills\u002Fship\u002FSKILL.md",[],[71,76,77,78,79,80,81,82,83,84,85,86,87,88,89],"Bash","l2","id","backlog","fix","edit","write","read","reviewer","grep","worktree","pr","tdd","review",{"skillNo":91,"name":92,"nameZh":6,"description":93,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":101,"publishedAt":38},"01M6KSVCP8NB3095MY8M2BR235","pr-review","对指定 PR 跑自动分级六维度 review（默认）；或 --check 模式验证上一轮 findings 是否修复 + 抓回归。按 diff 净增删行数自动分配 2\u002F3\u002F6 reviewer agent 并行（\u003C 200 行不派发，主 agent 自审）；主 agent 做根因聚类 + Cx 分级 + 修复分流建议，不自动 fix。","https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.codex\u002Fskills\u002Fpr-review\u002FSKILL.md",[96],{"categoryNo":97,"categoryName":98,"taxonomyType":99,"taxonomyTypeDesc":100},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[87,102,89,85,76,103,84,104,80,83],"pr review","check","diff",{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"publishedAt":38},"01M6KSV8P4AYPNHCDR1T7B9YN5","issues","GitHub Issues + Project v2 #3 项目管理单源技能。Part A：epic 拆解 + wave 实施顺序评论（找子任务 → blocked-by DAG → wave 1-4 容量装箱排序：每 wave ≤4、pri 优先、wave 内切并行组\u002F串行链、OPEN 重排、已完成与超窗单列 → 只追加 epic 评论；不写 Project 字段、不改 epic body）。Part B：issue\u002FPR 原子操作（建\u002F改 backlog issue、area\u002Ftype\u002Fpri label、PR 双轴状态 label 流转、统一 PR 评论格式 + 冲突预检\u002FCI watch 跟进，ship\u002Ffix 共用）。非 epic issue 号 → 查代码判状态（只判不修，建议 \u002Fship 或 close）。当用户要整理 epic 排 wave、建\u002F改 backlog issue、贴 label、切 PR 状态、给 PR 留评论、核一个 issue 是否还成立时使用。","https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.codex\u002Fskills\u002Fissues\u002FSKILL.md",[],[107,112,113,114,115,116,117,118,119,29,76,36,80,83,120,121,122,87,123,79],"项目管理","create-issue","labels","整理","EPIC","close","ci watch","type","open","ci","create","dag",{"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":38},"01M6KSV7CX440Q82S9VXQJH02G","git-worktree","Git Worktree 项目约定（编号、基准分支、权限兼容、删除安全）。","https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.codex\u002Fskills\u002Fgit-worktree\u002FSKILL.md",[],[131,86,132,133],"git","git worktree","安全",{"skillNo":135,"name":80,"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":38},"01M6KSV5JB0AR3EJ9V2H4F2MN2","问题诊断与修复: 验证+根因+复杂度分级+修复方案+backlog登记。当用户说'这个问题存在吗''帮我分析这个bug''诊断一下这个模块''修复这个问题'时触发。支持单条问题和多方审查报告批量输入。","https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.codex\u002Fskills\u002Ffix\u002FSKILL.md",[],[80,140,81,82,83,85,76],"文件",{"skillNo":142,"name":143,"nameZh":6,"description":144,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":145,"categories":146,"keywords":147,"publishedAt":38},"01M6KSV2S9ZK6BM2K1P2V3E4HK","speckit-taskstoissues","Convert existing tasks into actionable, dependency-ordered GitHub issues for the feature based on available design artifacts.","https:\u002F\u002Fgithub.com\u002Fghbvf\u002Fgocell\u002Fblob\u002Fdevelop\u002F.claude\u002Fskills\u002Fspeckit-taskstoissues\u002FSKILL.md",[],[148,18,21,149,22,23,25,26,150,27,29,151,152,34,107,153,154,35,36,155],"speckit taskstoissues","for","actionable","Existing","convert","artifacts","tasks","design"]