This is a comprehensive code review automation skill that orchestrates AI models (Claude, GPT-5, Qodo) alongside static analysis tools like CodeQL, SonarQube, and Semgrep. It walks through a full review workflow: triaging PRs by size and risk, running parallel security scans, routing to appropriate AI models based on complexity, and generating line-specific feedback with severity ratings and fix examples. The skill covers security (OWASP Top 10, secret detection), architecture patterns (microservices boundaries, SOLID principles), and performance regression detection. Use it when you want a structured framework for AI-assisted reviews rather than ad-hoc prompting. The routing logic and multi-tool coordination are solid, though you'll need to adapt the CI/CD integration and tool configs to your actual setup.
npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill code-review-ai-ai-review