# Code Review Reception

> Receives code review feedback through technical verification rather than performative agreement or blind implementation. Use this skill when receiving review comments, before implementing reviewer suggestions, when feedback is unclear, or when a suggestion may be wrong for this codebase. Do not use when/for authoring reviews of others' code or when branch-completion is only about landing already-agreed work.

- Skill: `jagreehal/code-review-reception` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add jagreehal/code-review-reception`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jagreehal/code-review-reception/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: jagreehal (https://skillmd.com/u/jagreehal)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/jagreehal/code-review-reception

---


# Code Review Reception

## Critical rules

- Verify before implementing. A suggestion is a hypothesis, not a command.
- Technical correctness before social comfort. No performative agreement.
- If any item is unclear: STOP; ask before implementing anything.
- Implement one item at a time; test each; confirm no regressions.
- Acknowledge correct feedback through action, not praise.
- Before drafting replies or pushback, read [references/examples.md](references/examples.md).

## Workflow

1. READ all feedback without reacting.
2. UNDERSTAND: restate each requirement (or ask).
3. VERIFY each suggestion against the codebase.
4. EVALUATE: technically sound for THIS codebase?
5. RESPOND: technical acknowledgment or reasoned pushback with evidence.
6. IMPLEMENT in order: blocking → simple → complex; test each fix.

## Resources

- [references/examples.md](references/examples.md) — unclear feedback, pushback, YAGNI, acknowledgment, rationalizations. Read when responding or unsure.

## Validation

Before responding:

- [ ] All feedback read; requirements restated or clarified
- [ ] Each suggestion checked against the codebase

Before claiming done:

- [ ] One item at a time; each tested; no regressions
- [ ] Correct items acknowledged factually; wrong ones pushed back with evidence

## Constraints

- Not for authoring reviews of others' code.
- Related: `verification-before-completion`, `tdd-workflow`, `debugging-methodology`, `critical-peer`.