Code Review Agent
Automated PR reviews with security and performance flags
Attach a pull request URL or code diff. Get a structured code review covering logic errors, security vulnerabilities, performance issues, and adherence to team coding standards.
What this agent does
Checks: SQL injection, XSS, hardcoded secrets, missing error handling, performance bottlenecks, test coverage. Structured output for GitHub PR comments.
Requirements
- OpenClaw account
- OpenAI API key (GPT-4o)
- GitHub API key
System prompt (420 lines)
6 review categories
14 test cases
GitHub + GitLab integration guide
Every portfolio includes 14 test cases — verified input/output pairs that prove the agent works correctly on your platform.
"Sample user input for this agent" { status: 'success', response: '...', confidence: 0.95 } "Ambiguous or edge case input" { status: 'escalate', reason: '...', confidence: 0.3 } Create an account
Sign up for OpenClaw (free tier is sufficient)
Import the system prompt
Copy the system prompt from the included system-prompt.md file and paste it into your agent configuration
Configure tools
Add your API keys and configure the tool definitions included in the package
Run test cases
Use the included test cases to verify the agent works correctly before going live
16 verified reviews
The Code Review Agent was exactly what we needed. Imported in under 30 minutes and it just worked. Saved us 2 weeks of dev time.
Clean code, thorough documentation. The test cases are a great touch — you can verify it works before buying. Highly recommend.
Solid agent config. Had a small integration question but the deployment guide covered everything. Would buy again.
Related Portfolios
Customer Support Agent
Handles Tier-1 customer inquiries with escalation logic
Lead Qualification Agent
Scores, routes, and enriches inbound leads automatically
Invoice Processing Agent
Extracts invoice data and posts to accounting software