菜单
Use when a planned or investigated change needs fresh verification evidence, layered gate proof, review truth, and an honest pass fail blocked verdict before entering cc-act.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | cc-check |
| version | 1.8.1 |
| description | Use when a planned or investigated change needs fresh verification evidence, layered gate proof, review truth, and an honest pass fail blocked verdict before entering cc-act. |
| triggers | ["验收这个需求","帮我确认是否完成","跑一下质量门","check this requirement","verify with evidence","can this ship","是不是可以进 cc-act"] |
| reads | ["PLAYBOOK.md","CHANGELOG.md","references/gate-contract.md","references/review-contract.md","assets/REPORT_CARD_TEMPLATE.json"] |
| writes | [{"path":"devflow/changes/<change-key>/review/report-card.json","durability":"durable","required":true}] |
| entry_gate | ["Read planning/design.md or planning/analysis.md, then planning/tasks.md, planning/task-manifest.json, and the latest runtime evidence before selecting a verdict.","Re-run fresh commands instead of inheriting cc-do narration.","If evidence is stale or missing, reset context and rebuild the verdict from canonical artifacts."] |
| exit_criteria | ["review/report-card.json records pass, fail, or blocked using fresh evidence, plus spec alignment and sync readiness.","Task-level review and requirement-level diff review are separated clearly.","The next step is unambiguous: cc-act, cc-do, cc-investigate, or cc-plan."] |
| reroutes | [{"when":"The implementation is incomplete, tests fail, or review findings require code changes.","target":"cc-do"},{"when":"The bug still lacks a trustworthy root cause or the investigation contract is disproven.","target":"cc-investigate"},{"when":"The plan assumptions, scope, or design contract are invalid.","target":"cc-plan"}] |
| recovery_modes | [{"name":"clean-room-reset","when":"The current verdict is contaminated by stale chat memory or old command output.","action":"Discard narrative memory, reload canonical artifacts, rerun fresh gates, and rebuild review/report-card.json from current evidence only."}] |
| tool_budget | {"read_files":8,"search_steps":4,"shell_commands":6} |
[PROTOCOL]: 变更时同步更新
version、CHANGELOG.md、相关模板/脚本引用，然后检查CLAUDE.md
cc-check 是 PDCA 里的 Check。
它负责把“应该好了”变成“证据表明它好了”。
它不是收尾话术器，也不是替 cc-do 涂绿。
NO PASS WITHOUT FRESH EVIDENCE
如果本轮没有重新读取契约、重新跑 gate、重新核对 review，就没有资格写 pass。
cc-check 的失败不是“你不够努力”，而是“现实还没有被证据证明为绿色”。
PLAYBOOK.mdCHANGELOG.mdreferences/gate-contract.mdreferences/review-contract.mdplanning/analysis.md如果代码还在继续变、任务还没收口，停下并回 cc-do。
先判断当前现实属于哪一类，再选 verdict：
| 现实状态 | 先读成什么 |
|---|---|
| 命令和 review 都齐了，只差最终裁决 | pass-candidate |
| 已有明确失败输出或 review finding | fail-candidate |
| 关键证据、review、环境、输入缺件 | blocked-candidate |
cc-check 的第一步不是跑更多形容词，而是先判“我是缺证据，还是有失败，还是已经足够过门”。
你必须按阶段推进，不能跳着给结论：
planning/design.md 或 planning/analysis.mdplanning/tasks.md、planning/task-manifest.jsoncc-do 叙述pass / fail / blockedreroutereview/report-card.json任何阶段发现“证据过期、边界矛盾、结论无法诚实成立”，都必须停下重建，而不是硬凑 pass。
cc-do; root-cause drift returns to cc-investigate; scope or design invalidation returns to cc-plan.cc-check 不是只看“测试是不是绿的”，而是至少看 4 层：
planning/task-manifest.json 里的 task proof 是否完整reviews.spec / reviews.code 是否与完成状态一致任何一层失真，都不能写 pass。
cc-check 的 requirement-level review 不能只写“diff 看过了”。至少要形成这些事实：
planning/tasks.md / task-manifest.json 抽取可执行项，逐项标 DONE / PARTIAL / NOT_DONE / CHANGED。这些事实写入 review.diffReview.details 或 review.findings。pass 只在 scope、completion、critical pass、doc staleness 都没有 blocking finding 时成立。
planning/design.md 或 planning/analysis.md，再读 planning/tasks.md、planning/task-manifest.json。spec/code review 证据，并能说明 expected spec delta 是否已被验证。planning/design.md 或 planning/analysis.mdplanning/tasks.md、planning/task-manifest.jsonpass 只在所有必要层都通过时成立fail 用于已有明确失败现实blocked 用于缺件、缺环境、缺 review、缺输入cc-docc-investigatecc-plancc-check 不允许“差不多算过”。结论只能从下面这张表里选：
| Verdict | 什么时候用 | reroute |
|---|---|---|
pass | 快速质量门通过，review gate 通过，当前证据足够支持“可以进入 cc-act” | none |
fail | 命令明确失败，或 review 给出明确未解决问题 | cc-do |
blocked | 缺 review 证据、缺必要前提、缺环境、缺输入，导致现在还不能诚实地下 pass / fail | cc-do、cc-investigate 或 cc-plan |
优先级永远是：
如果你已经知道实现方向错了、范围错了、设计前提失效了，blocked 不是答案，应该直接 reroute 到 cc-investigate 或 cc-plan。
如果你出现下面这些念头，说明 cc-check 已经跑偏：
cc-act”这些都不叫验证，叫叙述污染。回到 Reset Contract。
每条 evidence 至少要能回答这 3 件事：
不要写：
要写：
npm test exited 0, all targeted tests passednpm run lint exited 1, src/foo.ts:18 仍有 unused importspec review 缺失，当前无法确认 T003 是否满足 requirement如果系统是多层链路，不要只验最终按钮或最终命令。
至少要确认：
哪一层先坏，就把 blocking finding 写到那一层，不要把深层症状伪装成最终 verdict。
每个 finding 至少要写清：
critical / important / infohigh / medium / low，低置信不要伪装成 blockerruntime / task-review / diff-review / adversarial / docsfix-now / reroute-cc-do / reroute-cc-plan / reroute-cc-investigate / document-follow-up不能写“可能有问题”然后让接手者猜。要么证明，要么标成待验证 gap。
最小高质量 review/report-card.json 至少应该长这样：
{
"changeId": "REQ-123",
"verdict": "pass",
"overall": "pass",
"summary": "verdict=pass quick=3/3 strict=0/0 review=pass",
"quickGates": [],
"strictGates": [],
"review": {
"status": "pass",
"summary": "Task review and diff review both passed",
"details": "",
"taskReviews": { "status": "pass", "required": true, "summary": "all completed tasks carry spec/code proof", "reviewers": [], "findings": [] },
"diffReview": { "status": "pass", "required": true, "summary": "plan completion clean, no scope drift, no critical diff findings", "reviewers": [], "findings": [] },
"findings": []
},
"blockingFindings": [],
"reroute": "none",
"timestamp": "2026-04-15T12:00:00.000Z"
}
看完第一屏，下一位接手者应该立刻知道：
pass、fail 还是 blockedcc-do、cc-investigate 还是 cc-planreview/report-card.jsonCHANGELOG.mdreferences/gate-contract.mdreferences/review-contract.mdassets/REPORT_CARD_TEMPLATE.jsonscripts/run-quality-gates.shscripts/render-report-card.jsscripts/verify-gate.shpass / fail / blocked 的选择必须能从 verdict matrix 逐条解释。reroute 必须和阻塞原因一致，不能随手填默认值。cc-plan 或 cc-docc-actblocked 掩盖本该明确 reroute 的失败类型PLAYBOOK.mdCHANGELOG.mdreferences/gate-contract.mdUse when verified work must be shipped or handed off with a clear landing path: run simplify and required tests, create or update a PR, prepare a local handoff, close out merged work, sync docs, write release notes, and fold follow-ups back into backlog or roadmap.
咨询式证据化自我理解入口。Use when no usable profile exists, the user says they do not know what they want, or values/talents/interests/constraints need to be refreshed before advice.
重启人生 SKILL 的前门顾问与路由器。Use when the user asks for life/career direction, confusion triage, experiment/proof/integration routing, or starts without a usable ~/.restart-life profile.
周期收尾和 durable memory 写入层。Use when proof-card review needs evidence-backed PROFILE, COMPASS, PRINCIPLES, ROADMAP, BACKLOG, versions, and next-cycle handoff updates.
原则提炼层。Use when investigation has frozen a real pain/failure/conflict case and the user needs a reusable personal principle with case evidence, counterexample, review date, and life-act handoff.
私有人生路线图层。Use when the user asks for the next months/stages, after integration changes direction, or when ROADMAP/BACKLOG need evidence-backed refresh.