receiving-code-review
How to process code review feedback effectively. Understand before acting, verify suggestions technically, ask for clarification when needed. Complements the reviewer sub-agents.
- License 许可
- MIT
- 最近更新
- 2026-11-11
- Stars
- ★ 0
- Forks
- 4
01 前言 Front Matter
- name
- receiving-code-review
- description
- How to process code review feedback effectively. Understand before acting, verify suggestions technically, ask for clarification when needed. Complements the reviewer sub-agents.
- classification
- passive
03 SKILL.md 正文
Receiving Code Review
Guidance for processing review feedback — whether from human reviewers or sub-agent reviewers.
Principles
Understand before acting
- Read all feedback before making any changes
- Identify which comments are blocking vs advisory
- Group related comments that may have a common root cause
Verify suggestions technically
- Do not blindly implement reviewer suggestions — verify they are correct
- Run the suggested change through tests before accepting
- If a suggestion introduces a regression, flag it rather than accepting it
Ask for clarification
- If feedback is ambiguous, ask the reviewer to clarify rather than guessing
- If you disagree with a suggestion, explain why with evidence (test results, type errors, architectural constraints)
- Silence is not agreement — respond to every substantive comment
One concern at a time
- Address each piece of feedback in a separate commit where practical
- This makes review of the fixes easier and keeps the history clean
Anti-patterns
- Performative agreement: accepting all suggestions without evaluation
- Defensive dismissal: rejecting feedback without technical justification
- Scope creep: using review feedback as an excuse to refactor unrelated code
References
.agent/sub-agents/templates/— reviewer prompt templates.agent/directives/invoke-code-reviewers.md— when and how reviewers are invoked
04 同仓库更多 Skills
tsdoc
Write code-adjacent API documentation and the matching durable doc updates.
systematic-debugging
Structured debugging workflow. Use when encountering bugs, test failures, or unexpected behaviour. Reproduce, isolate, hypothesise, verify, fix, regression test.
start-right-thorough
Thorough grounding — full directive read, metacognition, and execution outline.
start-right-quick
Quick grounding at session start — read directives, memory, and check practice box.
receiving-code-review
How to process code review feedback effectively. Understand before acting, verify suggestions technically, ask for clarification when needed. Complements the reviewer sub-agents.
napkin
Maintain a per-repo napkin file at .agent/memory/napkin.md that tracks mistakes, corrections, and what works. Always active, every session, unconditionally. Read distilled.md and napkin.md before doing anything. Write to the napkin continuously as you work. Log your own mistakes, not just user corrections.