概要
The Multi-Agent Code Review skill for Claude Code automates the peer review process by running multiple specialized agents in parallel to audit your changes. It intelligently selects agents—such as security engineers, backend architects, or frontend specialists—based on your file diffs, then consolidates their findings into a single, actionable report. By applying a confidence-based filter (80% threshold) and automatic de-duplication, it eliminates noise and focuses your attention on critical vulnerabilities and maintainability issues before they reach production.
