Github Code Review

This deploys a coordinated swarm of specialized AI agents to review GitHub PRs, each focusing on a different domain like security vulnerabilities, performance bottlenecks, architectural patterns, or accessibility issues. Instead of running a single static analysis pass, you get parallel reviews from agents that can spot OWASP issues, suggest secure alternatives, analyze algorithm complexity, and enforce SOLID principles. The workflow integrates directly with GitHub CLI, so you can pipe PR data and diffs straight into the swarm and post results back as comments or review requests. The appeal here is coverage depth without manual coordination, though you'll want to tune which agents run based on your project's actual needs rather than throwing all five at every PR.

Install to Claude Code

npx -y skills add spencermarx/open-code-review --skill github-code-review --agent claude-code