Bringing Code Review to Claude Code

Claude
Claude Code introduces a multi-agent code review system to improve bug detection and address the growing bottleneck of code reviews.

Summary

Claude Code now features a comprehensive code review system, modeled after Anthropic's internal process, utilizing a team of agents to identify bugs that might be missed by human reviewers. This system addresses the increasing challenge of code review bottlenecks as developer output grows. Code Review dispatches agents to analyze pull requests (PRs) in parallel, verifying and ranking bugs by severity, and delivers a consolidated overview comment with in-line annotations. Testing shows significant improvements in bug detection, with 54% of PRs receiving substantive review comments compared to 16% previously. While more expensive than existing solutions, Code Review offers granular control over spend and usage through monthly caps, repository-level settings, and an analytics dashboard. It is currently available in research preview for Team and Enterprise plans.

(Source:Claude)