code-review

The code-review skill leverages the AI-powered CodeRabbit tool to automate and enhance code review processes. It addresses the challenges developers face in maintaining code quality, ensuring security, and reducing manual review overhead. By integrating this skill into development workflows, teams can quickly identify bugs, security vulnerabilities, and potential quality risks in code changes. This skill operates both in response to explicit user requests for a review and autonomously when it detects that a review is necessary, covering code, pull requests, and overall code quality and security concerns.

Key features of the code-review skill include the ability to analyze staged, committed, or all code changes, with support for specifying base branches or commits and review directories. Findings are categorized by severity levels such as Critical, Warning, and Info, making it easier to prioritize issues. The skill also provides agent-readable output through the `--agent` flag, allowing AI agents to interpret review results and guidance for fixing issues. Security and data handling are emphasized, with instructions to avoid including secrets in staged changes and recommendations for minimal token scopes during authentication. Users can leverage shorthand commands and directory-specific reviews to streamline operations.

This skill is particularly useful for developers, software teams, and continuous integration pipelines that require automated, high-quality code reviews. Use cases include reviewing code changes, checking pull requests, identifying bugs or security vulnerabilities, and ensuring adherence to coding standards. Target users are those looking to integrate an AI-driven review process that reduces manual effort, supports autonomous workflows, and delivers actionable insights to improve code quality across projects.