Loading...
Loading...
Use this after receiving code review feedback and before implementing suggestions, especially when feedback is unclear or technically questionable—requires technical rigor and verification, not perfunctory agreement or blind execution
npx skill4agent add jnmetacode/superpowers-zh receiving-code-reviewWhen receiving code review feedback:
1. Read: Read the feedback fully, don't rush to react
2. Understand: Paraphrase the requirement in your own words (or ask questions)
3. Verify: Check against the actual state of the codebase
4. Evaluate: Is this technically reasonable for this codebase?
5. Respond: Provide technical confirmation or a well-reasoned counterargument
6. Implement: Do one item at a time, test each individuallyIf any item is unclear:
Stop—don't implement anything yet
Ask for clarification on the unclear items
Why: Items may be interconnected. Partial understanding = incorrect implementation.Partner: "Fix items 1-6"
You understand 1, 2, 3, 6. Unsure about 4, 5.
❌ Wrong approach: Implement 1, 2, 3, 6 first, ask about 4,5 later
✅ Correct approach: "I understand items 1, 2, 3, 6. I need clarification on items 4 and 5 before proceeding."Before implementing:
1. Check: Is this technically correct for this codebase?
2. Check: Will it break existing functionality?
3. Check: Is there a reason the current implementation is written this way?
4. Check: Does it work across all platforms/versions?
5. Check: Does the reviewer have full context?
If the suggestion seems incorrect:
Push back with technical reasons
If verification isn't easy:
State the situation: "I can't verify this without [X]. Should I [investigate/ask/proceed with other tasks]?"
If it conflicts with previous partner decisions:
Stop and discuss with your partner firstIf the reviewer suggests "implementing it properly":
Grep the codebase for actual usage
If no one uses it: "This interface isn't being called. Should we delete it (YAGNI)?"
If someone uses it: Implement it properlyFor feedback with multiple items:
1. Clarify all unclear items first
2. Then implement in this order:
- Blocking issues (crashes, security)
- Simple fixes (spelling, imports)
- Complex fixes (refactoring, logic)
3. Test each fix individually
4. Verify no regressions✅ "Fixed. [Briefly describe what was changed]"
✅ "Good catch—[specific issue]. Fixed at [location]."
✅ [Fix directly and reflect it in the code]
❌ "You're absolutely right!"
❌ "Great point!"
❌ "Thanks for catching this!"
❌ "Thank you for [anything]"
❌ Any expression of gratitude✅ "You're right—I checked [X], and indeed [Y]. Implementing now."
✅ "After verification, I confirm you're correct. My initial understanding was wrong because [reason]. Fixing now."
❌ Long apologies
❌ Defend your pushback
❌ Overexplain| Mistake | Correction |
|---|---|
| Perfunctory agreement | Paraphrase requirements or take direct action |
| Blind implementation | Verify against the codebase first |
| Batch implementation without testing | Do one item at a time, test individually |
| Assuming reviewers are always right | Check if it will break existing functionality |
| Avoiding pushback | Technical correctness > social comfort |
| Starting implementation with partial understanding | Clarify all items first |
| Proceeding without verification | State limitations and request guidance |
Reviewer: "Delete legacy code"
❌ "You're absolutely right! Let me delete it..."Reviewer: "Delete legacy code"
✅ "Checked... The build target is 10.15+, and this API requires 13+. Backward compatibility requires keeping the legacy code. The current implementation has an incorrect bundle ID—should we fix it or drop pre-13 support?"Reviewer: "Implement comprehensive metrics tracking including database, date filtering, CSV export"
✅ "Grep'd the codebase—this interface isn't called anywhere. Should we delete it (YAGNI)? Or is there a call I'm missing?"Partner: "Fix items 1-6"
You understand 1, 2, 3, 6. Unsure about 4, 5.
✅ "I understand items 1, 2, 3, 6. I need clarification on items 4 and 5 before proceeding."gh api repos/{owner}/{repo}/pulls/{pr}/comments/{id}/replies