Practice Lesson 3 of 3

Team Operations

What this lesson teaches

Using Claude Code in a team requires clear communication and good practices. This lesson covers PRs, code reviews, and collaborative AI workflows.

Writing Good PRs

A PR should explain what and why, not just what changed:

## Summary
Add rate limiting to the API to prevent abuse.

## Changes
- Added express-rate-limit middleware
- Configured 100 requests per 15 minutes per IP
- Added rate limit headers to responses

## Testing
- Verified rate limiting triggers after 100 requests
- Confirmed headers are present in responses

## Related
- Plan: docs/plans/20240115_rate_limiting.md
- Issue: #42

Tip: Claude can help write PR descriptions. Ask it to summarize the changes you made together.

Receiving Reviews

When you get feedback:

  • Don't take it personally — Reviews are about code, not you
  • Ask for clarification — "Can you explain why this approach is better?"
  • Use Claude to help — "The reviewer suggested X. How would you implement that?"

Giving Reviews

When reviewing others' code:

  • Be specific — "This could cause a race condition" not "This seems wrong"
  • Suggest solutions — Don't just point out problems
  • Acknowledge good work — Positive feedback matters too

AI Collaboration Workflow

When working with AI in a team:

Be transparent

If Claude wrote significant code, mention it. Your team should know the context.

Own the output

You're responsible for code you submit, regardless of who (or what) wrote it. Review everything.

Share learnings

Good prompts and workflows benefit the whole team. Document what works.

Important: Never commit code you don't understand. If Claude generated something complex, make sure you can explain and maintain it.

Key Takeaways

  • PRs explain what, why, and how to test
  • Reviews are about code quality, not personal criticism
  • Be transparent about AI assistance
  • Own and understand all code you commit