claude-collab/examples/code-review/README.md

Claude Collab - The AI collaboration framework that prevents echo chambers

# Claude-Collab Code Review Example

This example demonstrates how Claude-Collab can be used for collaborative code reviews, simulating a GitHub pull request review process with multiple AI agents providing diverse perspectives.

## What it Shows

- **Multi-perspective Review**: Three specialized reviewers (performance engineer, security expert, UI specialist) examine code changes
- **Security Issue Detection**: Agents identify critical vulnerabilities (MD5 hashing, insecure tokens)
- **Collaborative Problem Solving**: Reviewers discuss and reach consensus on solutions
- **Anti-Echo Chamber**: Diverse viewpoints ensure thorough code review

## The Scenario

A developer has submitted a PR to update the authentication system. The changes include:
- Switching from plaintext to MD5 password hashing (problematic!)
- Replacing JWT tokens with base64 encoding (insecure!)
- Adding in-memory session storage (scalability issue!)

## Running the Demo

```bash
node pr-review-demo.js
```

The demo will:
1. Auto-start the Claude-Collab server if needed
2. Register three specialized reviewers
3. Display the PR changes
4. Show collaborative review process
5. Demonstrate consensus building
6. Conclude with all reviewers requesting changes

## Key Learnings

1. **Diverse Expertise**: Different specialists catch different issues
   - Performance engineer: scalability concerns
   - Security expert: critical vulnerabilities
   - UI specialist: user experience impact

2. **Collaborative Discussion**: Agents build on each other's feedback

3. **Consensus Building**: Team reaches agreement on best practices

4. **Quality Improvement**: Multiple perspectives lead to better code

## Customization Ideas

- Add more reviewers with different specialties
- Modify the code changes to test different scenarios
- Implement automated fix suggestions
- Add integration with actual GitHub PRs

## Use Cases

- Automated PR reviews
- Security vulnerability detection
- Code quality assessment
- Team training simulations
- Best practices enforcement