技能 receiving-code-review
👁️

receiving-code-review

安全 🌐 网络访问⚙️ 外部命令

Evaluate code review feedback with technical rigor

也可从以下获取: DYAI2025,CodingCossack,obra,Cycleaddict,sickn33,ZhanlinCui,Cygnusfear

AI assistants often implement code review feedback without verification. This skill ensures technical evaluation of suggestions, asking clarifying questions and pushing back when feedback is unclear or technically incorrect.

支持: Claude Codex Code(CC)
📊 69 充足
1

下载技能 ZIP

2

在 Claude 中上传

前往 设置 → 功能 → 技能 → 上传技能

3

开启并开始使用

测试它

正在使用“receiving-code-review”。 Reviewer suggested removing legacy API code and adding proper metrics tracking

预期结果:

  • Checking legacy API usage - grep shows 3 call sites for backward compat (OS 10.15+)
  • Suggesting clarification: verify if metrics endpoint is actually called before implementing
  • Will clarify scope before implementing either change

安全审计

安全
v5 • 1/17/2026

Pure prompt-based skill containing only markdown documentation for AI behavior during code review feedback reception. No executable code, scripts, network calls, filesystem access, or external command execution. The static findings from the secondary scan are FALSE POSITIVES caused by the scanner misinterpreting markdown code fences (backticks) as shell backtick execution and documentation references as cryptographic or network threats.

2
已扫描文件
387
分析行数
2
发现项
5
审计总数
审计者: claude 查看审计历史 →

质量评分

38
架构
100
可维护性
83
内容
20
社区
100
安全
91
规范符合性

你能构建什么

Technical verification of suggestions

Ensure AI assistants verify code review feedback against codebase before implementing, preventing blind agreement

Consistent review handling

Maintain technical rigor when AI agents process external code review feedback from contributors

Quality feedback validation

Prevent AI assistants from implementing questionable suggestions without proper technical evaluation

试试这些提示

Basic review check
I am receiving code review feedback. Read all items completely. For each item: restate the requirement in my own words, verify against the codebase, and evaluate if technically sound. Ask clarification on unclear items before implementing anything.
External reviewer
External reviewer suggested: [suggestion]. Before implementing: check if this is technically correct for this codebase, if it breaks existing functionality, and if the reviewer understands the full context. Push back with technical reasoning if needed.
YAGNI verification
Reviewer suggests adding a feature. First grep the codebase to verify actual usage. If unused, suggest removal (YAGNI). If used, then implement properly. Report findings before proceeding.
Multi-item clarification
I received N items of feedback. Identify any unclear items. Do NOT implement partial changes. State what I understand and what needs clarification before touching any code.

最佳实践

  • Verify suggestions against actual codebase usage before implementing
  • Ask clarifying questions on any unclear feedback items before proceeding
  • Use technical reasoning when pushing back, not defensiveness

避免

  • Blindly implementing feedback without verification
  • Using performative agreement ('You are absolutely right')
  • Implementing partial items when some are unclear
  • Avoiding pushback on technically incorrect suggestions

常见问题

Is this skill compatible with all Claude models?
Yes, the skill works with Claude, Codex, and Claude Code. It provides behavioral guidelines for code review handling.
What happens if feedback is technically incorrect?
The skill guides you to push back with technical reasoning, referencing tests or codebase evidence to support your position.
Can this skill integrate with existing workflows?
Yes, it overlays on any code review process. Apply the guidelines before implementing feedback in your current workflow.
Does this skill access any files or data?
No, this is a prompt-based skill. It provides behavioral instructions only and does not access any files or systems.
What if I cannot verify a suggestion?
State your limitation clearly: 'I cannot verify without X. Should I investigate further, ask, or proceed?'
How is this different from standard code review handling?
It emphasizes technical verification over social performance, requiring evidence-based evaluation rather than blind agreement.