
receiving-code-review
✓ Official★ 4,933by microsoft · part of microsoft/fluidframework
Every review comment gets verified against the codebase before it's touched, then fixed and tested one at a time and pushed as a batch. Suggestions that don't hold up get factual technical pushback instead of automatic agreement.
WHEN YOUR AGENT SHOULD USE IT
A QUICK BOUNDARYUSE FOR
- Work through PR feedback item by item instead of batching untested changes.
- Push back with technical reasoning when a suggestion breaks something or adds functionality nobody asked for.
- Get clarification on every unclear review comment before implementing any of them.
- Fetch and read all PR comments and turn them into a tracked checklist.
This is the playbook your agent receives when the skill activates — you don't need to read it to use the skill, but it's here to audit before installing.
Code Review Reception
Overview
Code review requires technical evaluation, not emotional performance.
Core principle: Fetch feedback → Verify → Implement → Re-test → Push updates.
Announce at start: "I'm using the Nori Receiving Code Review skill to handle this feedback."
The Process
Step 0: Create Todo List
For multi-item feedback, use TodoWrite:
- [ ] Fetch and read all PR comments
- [ ] Clarify unclear items (if any)
- [ ] Fix item 1: [description]
- [ ] Fix item 2: [description]
...
- [ ] Run tests/lint/format
- [ ] Push updatesWhy: Prevents skipping items and provides visibility to user.
Step 1: Fetch PR Comments
Determine PR number from context:
- User mentions PR number: Use that
- Current branch: Run
gh pr view --json number -q .number
Fetch all comments:
# View all comments (review + general)
gh pr view [PR-NUMBER] --commentsRead completely before reacting.
Step 2: Understand and Clarify
Apply these checks to each item:
- Can I restate this requirement in my own words?
- Is this technically sound for THIS codebase?
- Does this break existing functionality?
- Is there a reason for the current implementation?
CRITICAL: If ANY item is unclear, STOP. Ask for clarification on ALL unclear items before implementing ANYTHING.
Example:
User: "Fix items 1-6"
You understand 1,2,3,6. Unclear on 4,5.
✅ "Understand 1,2,3,6. Need clarification on 4 and 5 before implementing."
❌ Implement 1,2,3,6 now, ask about 4,5 laterStep 3: Implement Changes
Follow implementation order:
- Blocking issues (breaks, security)
- Simple fixes (typos, imports)
- Complex fixes (refactoring, logic)
For each fix:
- Implement one at a time
- Test individually
- Commit individually (frequent commits)
YAGNI check: If reviewer suggests "implementing properly", grep for actual usage:
grep -r "endpointName" .If unused: "This endpoint isn't called. Remove it (YAGNI)?"
Step 4: Run Tests, Lint, and Format
Reference finishing-a-development-branch skill (Steps 1-2):
See .claude/skills/finishing-a-development-branch/SKILL.md
- Run tests:
npm test(or project equivalent)- If tests fail, fix before proceeding
- Run type checks:
npm run lint:*-types(if available)- If type errors, fix before proceeding
- Run formatter:
npm run format - Run linter:
npm run lint - Verify changes:
git diff --stat
Step 5: Push Updates
Push changes to PR:
git pushStep 6: Summary and Next Action
Report what changed:
"Code review feedback addressed:
- Fixed [item 1]: [brief description]
- Fixed [item 2]: [brief description] ...
Changes pushed to PR. Options:
- Done - PR is ready for re-review
- More feedback - Additional changes needed
- Show changes - Review diffs before marking done
Which would you like?"
Quick Reference Checklist
- Create TodoWrite for all feedback items
- Fetch PR comments (
gh pr view --comments) - Clarify ALL unclear items before implementing ANY
- Implement in order: blocking → simple → complex
- Test each fix individually
- Run tests, type checks, formatting, linting (finishing-a-development-branch)
- Push updates
- Summarize changes and ask for next action
Response Tone Guidelines
Forbidden:
- "You're absolutely right!" / "Great point!" / "Thanks for..." (performative agreement)
- Implementing before verifying against codebase
- Proceeding with unclear feedback
Required:
- Verify suggestions against codebase reality before implementing
- Push back with technical reasoning if suggestion breaks things or violates YAGNI
- Ask for clarification on ALL unclear items before implementing ANY items
- State fixes factually: "Fixed. [what changed]" or just show the code
YAGNI check: If reviewer suggests "implementing properly", grep for actual usage. If unused, ask: "This endpoint isn't called. Remove it (YAGNI)?"
When wrong after pushing back: "You were right - checked [X] and it does [Y]. Implementing now." No apology needed.
External reviewers: Check if technically correct for THIS codebase, works on all platforms, doesn't conflict with partner's prior decisions. If conflicts, discuss with partner first.
Signal discomfort pushing back: "Strange things are afoot at the Circle K"
Common Mistakes
| Mistake | Fix |
|---|---|
| Performative agreement | State requirement or just act |
| Blind implementation | Verify against codebase first |
| Batch without testing | One at a time, test each |
| Assuming reviewer is right | Check if breaks things |
| Avoiding pushback | Technical correctness > comfort |
| Partial implementation | Clarify all items first |
| Can't verify, proceed anyway | State limitation, ask for direction |
Red Flags
Never:
- Skip creating TodoWrite for multi-item feedback
- Implement without verifying against codebase
- Proceed with unclear feedback
- Skip tests/linting/formatting before pushing
Always:
- Read all feedback completely first
- Clarify unclear items before implementing
- Test each fix individually
- Run full verification before pushing
- Provide summary of changes to user
Install this skill from microsoft/fluidframework on GitHub. It only works when installed from that project, so don't copy the skill's file out on its own.
npx skills add microsoft/fluidframework --skill "receiving-code-review" --full-depthRun this in your project — your agent picks the skill up automatically.
BEFORE IT WILL WORK
1 FOR YOU- 01Have `gh` authenticated for the PR
Licensed under MIT— you can use, modify, and redistribute it under that license's terms.
View the full license file on GitHub →