Token导航 LogoToken导航

Agent Skill · 研究检索

ccpm-code-review(ccpm 代码审查)

ccpm-code-review 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

ccpm-code-review 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证为 MIT,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 70,604。

总安装 210周安装 9GitHub Stars 9下载 73核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:ccpm-code-review(ccpm 代码审查)
来源仓库:https://github.com/duongdev/ccpm
仓库路径:skills/ccpm-code-review
安装命令:
npx skills add https://github.com/duongdev/ccpm --skill ccpm-code-review
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/duongdev/ccpm --skill ccpm-code-review

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

该技能可确保: ✅ 没有虚假的完工声明 ✅ “完成”之前需要的证据 ✅ 实施质量关卡(4 步验证) ✅ 技术严谨性高于社交舒适度 ✅ 与 CCPM 自然工作流程集成 ✅ 出现故障时进行系统调试 理念:完成前验证,证据重于声明,质量重于速度。

主要特点: 尝试完成时自动激活 实施 4 个验证门 与 /ccpm:verify 集成 和/ccpm:完成 建议/ccpm:验证 对于失败 与外部系统安全配合进行确认 资料来源:改编自 claudekit-skills/code-review 许可证:麻省理工学院 CCPM 集成:/ccpm:verify, /ccpm:完成, /ccpm:验证, 质量门钩 每周安装量 9 存储库 duongdev/ccpm GitHub 之星 9 第一次看到 2026 年 1 月 24 日 安全审计 Gen Agent Trust Hub 通行证 套接字通行证 斯尼克通行证

来源仓库内路径:skills/ccpm-code-review

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/duongdev/ccpm --skill ccpm-code-review;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

ccpm-code-review 怎么安装?

在终端执行 npx skills add https://github.com/duongdev/ccpm --skill ccpm-code-review。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

ccpm-code-review 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

ccpm-code-review 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

ccpm-code-review 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

ccpm-code-review 用的人多吗?

公开数据是累计安装 210 次,近一周安装 9 次,下载 73 次,GitHub 9 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

ccpm-code-review 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
ccpm-code-review

ccpm 代码审查 · 本页

2109需要联网MIT通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

CCPM Code Review

Structured code review workflow integrated with CCPM's Linear-based project management system. Enforces "no completion claims without verification evidence" principle.

When to Use

This skill auto-activates when:

  • User says "done", "complete", "finished", "ready to merge"
  • Running /ccpm:verify command (natural workflow - recommended)
  • Running /ccpm:done command (includes pre-flight verification)
  • Running /ccpm:verify command (advanced)
  • Before updating Linear task status to "Done"
  • Before syncing Jira status
  • Before creating GitHub/BitBucket PR
  • Before sending Slack completion notifications

CCPM Verification Workflow

CCPM provides a streamlined 3-step verification process:

Step 1: Quality Checks (Automated)

Run linting, tests, and build checks to ensure technical correctness:

/ccpm:verify [issue-id]  # Auto-detects issue from git branch

What it checks:

  • ✅ Linting passes (no style errors)
  • ✅ Tests pass (all unit and integration tests)
  • ✅ Build succeeds (no compilation errors)
  • ✅ Checklist complete (100% of implementation items)

If checks fail: Command automatically suggests /ccpm:verify to debug systematically.

Step 2: Agent Code Review

After quality checks pass, agent review analyzes:

  • ✅ Code quality and best practices
  • ✅ Security vulnerabilities
  • ✅ Performance implications
  • ✅ Requirement fulfillment
  • ✅ Regression risks

Step 3: Final Confirmation

Four verification gates must pass:

  1. Tests Pass ✅ - Zero failures
  2. Build Succeeds ✅ - Exit status 0
  3. Checklist Complete ✅ - 100% checked
  4. No Blockers ✅ - No unresolved blockers in Linear

Only after all gates pass can task proceed to finalization with /ccpm:done.

Core Principles

1. Technical Correctness Over Social Comfort

Forbidden performative agreement:

  • ❌ "Great point!"
  • ❌ "You're absolutely right!"
  • ❌ "That makes total sense!"

Required rigorous verification:

  • ✅ "Let me verify that assumption"
  • ✅ "I'll test this approach first"
  • ✅ "Here's the evidence: [test output]"

2. No Implementation Before Verification

When receiving feedback:

  1. First: Verify the feedback is technically correct
  2. Then: Implement the change
  3. Never: Blindly implement without understanding

3. NO COMPLETION CLAIMS WITHOUT EVIDENCE

Required evidence before any "done" claim:

  • ✅ Tests: All passing (screenshot or CI link)
  • ✅ Build: Exit status 0 (no errors)
  • ✅ Linear checklist: 100% complete
  • ✅ No unresolved blockers in Linear comments

Integration with CCPM Commands

With /ccpm:verify (Natural Workflow - Recommended)

This skill enforces verification during the streamlined verify command:

# User runs natural verification command
/ccpm:verify

# Or with explicit issue ID
/ccpm:verify AUTH-123

Flow:

User: "I'm done, let me verify"

Claude: [ccpm-code-review activates]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔍 Smart Verify Command
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

📋 Issue: AUTH-123 - Add user authentication
📊 Status: In Progress
📋 Checklist: 100% (5/5 items)

Verification Flow:
──────────────────
1. Quality Checks (linting, tests, build)
2. Final Verification (code review, security)

═══════════════════════════════════════
Step 1/2: Running Quality Checks
═══════════════════════════════════════

🔍 Running linting...
✅ All files pass linting

🧪 Running tests...
✅ All tests passed (28/28)

🏗️  Running build...
✅ Build successful

📊 Quality Check Results:
  ✅ Linting
  ✅ Tests
  ✅ Build

═══════════════════════════════════════
Step 2/2: Running Final Verification
═══════════════════════════════════════

[Code reviewer agent analyzes changes...]

✅ All requirements met
✅ Code quality standards met
✅ Security best practices followed
✅ Performance acceptable

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
✅ All Verification Complete!
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

All verifications passed! Ready to finalize.

💡 What's Next?
⭐ Recommended: /ccpm:done AUTH-123

If verification fails:

❌ Quality Checks Failed

📊 Quality Check Results:
  ✅ Linting
  ❌ Tests (3 failures)
  ✅ Build

To debug and fix issues:
  /ccpm:verify AUTH-123

Then run verification again:
  /ccpm:verify AUTH-123

With /ccpm:done (Pre-Flight Verification)

This skill ensures quality before finalization:

# User attempts to finalize task
/ccpm:done

# Or with explicit issue ID
/ccpm:done AUTH-123

Pre-flight checks enforced by ccpm-code-review:

User: "/ccpm:done AUTH-123"

Claude: [ccpm-code-review activates for pre-flight]

✅ All pre-flight checks passed!

Checking completion readiness:

□ Tests passing?
  → ✅ All 52 tests passed

□ Build successful?
  → ✅ Build completed (exit 0)

□ Linear checklist complete?
  → ✅ 7/7 items checked

□ No blockers?
  → ✅ No blockers found

VERIFICATION COMPLETE ✅

📝 Creating GitHub Pull Request...
✅ Pull Request created

[Prompts for Jira/Slack updates with confirmation]

✅ Linear issue updated to Done

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🎉 Task Finalized: AUTH-123
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

If pre-flight fails:

⛔ Cannot Finalize: Checklist Incomplete

Progress: 80% (4/5 completed)

❌ Remaining Items:
  - [ ] Write integration tests

🔧 Actions Required:
1. Complete remaining checklist items
2. Update checklist: /ccpm:sync AUTH-123
3. Then run: /ccpm:done AUTH-123

Advanced: With /ccpm:verify

For manual verification step (advanced users):

Claude runs /ccpm:verify AUTH-123

[ccpm-code-review activates during verification]

Checking verification evidence:

□ Tests passing?
  → ✅ All 47 tests passed

□ Build successful?
  → ✅ Build completed with exit code 0

□ Linear checklist complete?
  → ❌ 3 of 5 items checked

□ No blockers?
  → ⚠️ 1 blocker found: "Redis connection failing in staging"

VERIFICATION FAILED:
- Linear checklist incomplete (3/5)
- Blocker unresolved: Redis connection issue

Cannot proceed with completion until:
1. Complete remaining checklist items
2. Resolve Redis blocker or document as out-of-scope

Would you like to:
A. Update checklist now
B. Investigate Redis blocker
C. Update task scope to exclude Redis

Receiving Feedback Protocol

When PR Reviewer Comments

Scenario: Reviewer says "Add error handling to login function"

❌ Wrong approach (performative):

"Great point! I'll add error handling right away."
[Immediately adds try-catch without understanding]

✅ Correct approach (verification-first):

Claude: [ccpm-code-review activates]

Let me verify this feedback:

1. Checking current error handling...

// login.ts async function login(email, password) { const user = await db.findUser(email); // ← No error handling return generateToken(user); // ← No error handling }


1. Potential issues:
  - Database query could fail (network, timeout)
  - user might be null (not found)
  - generateToken could throw (invalid data)
2. Verification: Feedback is VALID
  - Missing error handling confirmed
  - Could cause uncaught exceptions
3. Implementation approach: `async function login(email, password) {try {const user = await db.findUser(email); if (!user) {throw new UnauthorizedError('Invalid credentials');} return generateToken(user);} catch (error) {if (error instanceof DatabaseError) {logger.error('Database error during login', error); throw new ServiceUnavailableError();} throw error;}}`
4. Testing the fix:
  - Add test: login with network error → returns 503
  - Add test: login with invalid user → returns 401
  - Verify existing tests still pass
5. Update Linear:
  - Add comment: "PR feedback: Added error handling"
  - Link to new commit

Implementing fix now...

When Feedback Seems Wrong

Scenario: Reviewer says "Use var instead of const"

❌ Wrong approach:


"Okay, I'll change it!" [Blindly makes change]

✅ Correct approach:


Claude: [ccpm-code-review activates]

Let me verify this feedback:

1. Current code uses `const` (ES6+)
2. Feedback suggests `var` (ES5)
3. Technical analysis:
  - `const`: Block-scoped, immutable binding, prevents reassignment bugs
  - `var`: Function-scoped, allows reassignment, hoisting issues
4. Project standards check:
  - .eslintrc: Enforces `const`/`let`, disallows `var`
  - Existing codebase: 100% uses `const`/`let`
5. Verification: Feedback appears INCORRECT
  - Contradicts project standards
  - Would cause linting failures
  - Downgrades code quality
6. Response to reviewer: "I checked our ESLint config and existing codebase. We enforce `const`/`let` and disallow `var` for block-scoping safety. Changing to `var` would: Did you mean something else, or is there a specific reason for `var` here?"
  - Fail CI linting
  - Contradict project standards
  - Introduce potential hoisting bugs

Waiting for reviewer clarification before making changes...

Verification Gates

Gate 1: Tests Must Pass

Requirement: Zero test failures


✅ PASS: 52 tests, 0 failures ❌ FAIL: 48 tests passed, 4 failures

If failures exist:

1. Run /ccpm:verify to debug
2. Update Linear with findings
3. Re-run verification after fixes
4. Only proceed when 0 failures

Gate 2: Build Must Succeed

Requirement: Exit status 0


✅ PASS: Build completed successfully (exit 0) ❌ FAIL: Build failed with 3 TypeScript errors (exit 1)

If build fails:

1. Review error messages
2. Fix compilation errors
3. Re-run build
4. Only proceed when exit 0

Gate 3: Linear Checklist Must Be Complete

Requirement: 100% of checklist items checked


✅ PASS: 7/7 checklist items complete ❌ FAIL: 5/7 checklist items complete

Incomplete items:

- Add integration tests
- Update API documentation

Action required:

1. Complete remaining items, OR
2. Update task scope (with /ccpm:plan)
3. Mark items as out-of-scope explicitly
4. Only proceed when justified

Gate 4: No Unresolved Blockers

Requirement: Zero blocker comments in Linear


✅ PASS: No blockers found ❌ FAIL: 2 blockers in Linear comments

Blockers:

1. "Database migration failing in staging"
2. "Awaiting security team approval"

Action required:

1. Resolve each blocker, OR
2. Document why blocker doesn't apply
3. Get explicit approval to proceed despite blocker
4. Only proceed when all blockers addressed

Integration with CCPM Safety Rules

Works with external-system-safety Skill

This skill focuses on verification, external-system-safety focuses on confirmation.

Division of responsibility:

ConcernSkillCheck
Are tests passing?ccpm-code-review✓
Is build successful?ccpm-code-review✓
Is checklist complete?ccpm-code-review✓
User confirms Jira update?external-system-safety✓
User confirms Slack post?external-system-safety✓

Together they ensure:

  1. Quality standards met (ccpm-code-review)
  2. User approves external writes (external-system-safety)

References SAFETY_RULES.md

All external writes follow ../commands/SAFETY_RULES.md:

⛔ Require confirmation:

  • Jira status updates
  • Confluence page edits
  • BitBucket PR creation
  • Slack notifications

✅ Always allowed:

  • Linear status updates (internal tracking)
  • Local file modifications
  • Read operations

Examples

Example 1: Successful Completion Flow (Natural Workflow)


User: "Done with AUTH-123, ready to ship!"

Claude: [ccpm-code-review activates]

Let me run verification first:

Running /ccpm:verify AUTH-123...

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 🔍 Smart Verify Command ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

📋 Checklist: 100% (7/7 items)

═══════════════════════════════════════ Step 1/2: Running Quality Checks ═══════════════════════════════════════

✅ Tests: 52/52 passing ✅ Build: Successful (exit 0) ✅ Linting: No errors

═══════════════════════════════════════ Step 2/2: Running Final Verification ═══════════════════════════════════════

✅ Code review: Passed ✅ Security audit: Passed ✅ Requirements met: Passed

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ ✅ All Verification Complete! ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

💡 What's Next? ⭐ Recommended: /ccpm:done AUTH-123

Ready to finalize! Shall I run /ccpm:done now?

Example 2: Failed Verification (Natural Workflow)


User: "I'm done, let's merge"

Claude: [ccpm-code-review activates]

Running /ccpm:verify...

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 🔍 Smart Verify Command ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

═══════════════════════════════════════ Step 1/2: Running Quality Checks ═══════════════════════════════════════

🧪 Running tests... ❌ 3 tests failing:

- test/auth/login.test.ts: "should handle invalid credentials"
- test/auth/jwt.test.ts: "should refresh expired tokens"
- test/api/protected.test.ts: "should reject missing tokens"

🏗️ Running build... ❌ TypeScript error: src/auth/jwt.ts:45:12 - error TS2345

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ ❌ Verification Failed ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

To debug and fix issues: /ccpm:verify AUTH-123

Then run verification again: /ccpm:verify AUTH-123

Would you like me to start debugging these issues?

Example 3: Incomplete Checklist (Pre-Flight with /ccpm:done)


User: "/ccpm:done AUTH-456"

Claude: [ccpm-code-review activates for pre-flight]

✅ All git pre-flight checks passed!

Checking completion readiness:

✅ Tests passing (41/41) ✅ Build successful ❌ Linear checklist incomplete (4/6): ✓ Implement JWT generation ✓ Add login endpoint ✓ Add logout endpoint ✓ Write unit tests ✗ Add integration tests ✗ Update API documentation

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ ⛔ Cannot Finalize: Checklist Incomplete ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Progress: 67% (4/6 completed)

🔧 Actions Required:

1. Complete remaining checklist items
2. Update checklist: /ccpm:sync AUTH-456
3. Then run: /ccpm:done AUTH-456

Or update scope with /ccpm:plan AUTH-456 "mark items as Phase 2"

Which would you like to do?

Example 4: Complete Natural Workflow (End-to-End)


User: "I finished the authentication feature"

Claude: [ccpm-code-review guides through workflow]

Great! Let's verify your work before finalizing.

Step 1: Verification ──────────────────── /ccpm:verify

✅ All quality checks passed ✅ Code review complete ✅ Security validated

Step 2: Finalization ──────────────────── /ccpm:done

✅ Git pre-flight checks passed ✅ Checklist complete (5/5) ✅ No blockers 📝 GitHub PR created ✅ Linear updated to Done

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 🎉 Task Finalized: AUTH-456 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

All done! Task completed successfully. 🎊

Verification Checklist Template

Before claiming "done":

  • [ ] Tests executed

- [ ] All tests passing (0 failures) - [ ] Coverage meets requirements - [ ] New tests added for new code

  • [ ] Build verified

- [ ] Build successful (exit 0) - [ ] No compilation errors - [ ] No linting errors

  • [ ] Linear task complete

- [ ] All checklist items checked - [ ] No unresolved blockers - [ ] Work summary added

  • [ ] Code quality

- [ ] Code reviewed (by human or code-reviewer agent) - [ ] Security checked (if applicable) - [ ] Performance acceptable

  • [ ] Documentation

- [ ] Code comments added where needed - [ ] API docs updated (if API changed) - [ ] README updated (if user-facing)

Only after ALL checked:

  • [ ] Ready for /ccpm:done

Integration with Other CCPM Skills

Works alongside:

  • external-system-safety: Enforces confirmation for external writes (Jira/Slack)
  • pm-workflow-guide: Suggests /ccpm:verify and /ccpm:done at right time
  • ccpm-debugging: Invoked via /ccpm:verify when checks fail
  • sequential-thinking: For complex verification scenarios

Example combined activation:


User: "Ready to merge AUTH-123" ↓ ccpm-code-review → Suggests /ccpm:verify first ↓ /ccpm:verify → Runs quality checks + agent review ↓ [If gates pass] ↓ Suggests /ccpm:done ↓ /ccpm:done → Pre-flight checks + PR creation ↓ external-system-safety → Confirms Jira/Slack writes ↓ [If user confirms] ↓ Complete! ✅

Natural Workflow Commands

CCPM provides streamlined commands for the complete verification and finalization workflow:

CommandPurposeAuto-detects Issue
/ccpm:verifyQuality checks + agent review✅ From git branch
/ccpm:donePre-flight + PR + finalize✅ From git branch
/ccpm:verifyDebug failed checks❌ Explicit ID required

Recommended workflow:

# 1. Complete implementation
/ccpm:work

# 2. Commit changes
/ccpm:commit

# 3. Verify quality (this skill activates)
/ccpm:verify

# 4. Finalize task (this skill activates for pre-flight)
/ccpm:done

Summary

This skill ensures:

  • ✅ No false completion claims
  • ✅ Evidence required before "done"
  • ✅ Quality gates enforced (4-step validation)
  • ✅ Technical rigor over social comfort
  • ✅ Integration with CCPM natural workflow
  • ✅ Systematic debugging when failures occur

Philosophy: Verification before completion, evidence over claims, quality over speed.

Key Features:

  • Auto-activates on completion attempts
  • Enforces 4 verification gates
  • Integrates with /ccpm:verify and /ccpm:done
  • Suggests /ccpm:verify for failures
  • Works with external-system-safety for confirmations

Source: Adapted from claudekit-skills/code-review License: MIT CCPM Integration: /ccpm:verify, /ccpm:done, /ccpm:verify, quality-gate hook

平台分布

Claude Code

29.26%
按下载量换算21

OpenCode

21.08%
按下载量换算15

Cursor

18.1%
按下载量换算13

Gemini CLI

12.16%
按下载量换算9

kilo

7.39%
按下载量换算5

windsurf

3.54%
按下载量换算3

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。

继续浏览同类 Skills