| name | apply-review-feedback |
| description | Triage and apply code review feedback with technical verification. Use when a human, reviewer subagent, CI, or adversarial reviewer returns findings or requested changes. |
| version | 0.1.0 |
| author | George-RD |
| x-compatibility | jcode-primary, claude-code-adaptable, opencode-adaptable |
Apply Review Feedback
Use this skill when receiving review feedback. Treat feedback as claims to verify, not orders to blindly follow.
Core principle: understand, verify, classify, then fix one item at a time with gates.
Workflow
- Read all feedback before editing.
- Classify each item: accept, reject with evidence, clarify, defer, or already fixed.
- If any item is unclear and may affect other items, ask or investigate before partial implementation.
- Verify claims against code, tests, requirements, and repo workflow.
- Fix accepted Critical items first, then Important, then opportunistic Minor items.
- Run targeted gates after each fix cluster and full gates before finalizing.
- Report what changed, what was rejected or deferred, and why.
Context funnel
Read references only when needed:
references/feedback-triage.md: classification workflow and response templates.references/yagni-check.md: when reviewers ask for professional but unused features.references/pushback-patterns.md: how to reject incorrect feedback with evidence.scripts/validate.sh: behavioral contract checks for this skill.
Rules
- Avoid performative agreement. Be concise and technical.
- Do not implement unclear multi-item feedback partially if dependencies are possible.
- Do not add scope because a reviewer said "properly" or "production ready". Check actual usage and requirements.
- If feedback conflicts with user decisions or project specs, stop and surface the conflict.
