Anthropic Announces Agentic Code Reviews for Claude Code
Anthropic's new team-based review system that dispatches targeted agents on every PR to catch the bugs that skims miss, built for depth, not speed.
Anthropic just announced its new agentic Code Review system, designed to work seamlessly with Claude Code and GitHub Pull Requests. Not to be confused with their previous and underwhelming Claude Code GitHub Action. This new GitHub App dispatches a "deep, multi-agent review process," with each agent responsible for a different class of issue. This approach is "more thorough (and more expensive)" than the existing GitHub Action. They estimate a PR review may average a cost of $15-25, scaling with PR size and task complexity.
According to the article, they've been running Code Review internally for their PRs for months:
We've been running Code Review internally for months: on large PRs (over 1,000 lines changed), 84% get findings, averaging 7.5 issues. On small PRs under 50 lines, that drops to 31%, averaging 0.5 issues. Engineers largely agree with what it surfaces: less than 1% of findings are marked incorrect.
Introduces the REVIEW.md repository file
In addition to reviewing your repository's root CLAUDE.md, you can specify a REVIEW.md to provide review-specific guidance without polluting your general coding session instructions.
What's next?
This seems like a promising feature. Code Review is now available as a research preview for customers on the Team and Enterprise plans. Check out the announcement video here: