Review

This is a code review skill that runs static analysis on your codebase and generates actionable feedback. It uses tree-sitter for parsing, runs language-specific linters, and outputs structured reviews with severity levels and suggested fixes. You'd reach for this when you want automated feedback before committing or as part of a CI pipeline. The interesting bit is that it's designed to work with Claude's agent framework, so it can iteratively improve code based on its own suggestions. The skill definition is minimal though, so you'll need to dig into the implementation to see exactly which languages and linters it supports out of the box.