Claude Code Ships Multi-Agent Code Review — and Anthropic Says It Already Tripled Its Own PR Throughput
Anthropic publicly launched a code review feature for Claude Code that deploys multiple agents to hunt bugs across pull requests, ranking them by severity. Internally, the tool boosted Anthropic's PR review rate from 16% to 54%.
Anthropic released a new Code Review capability for Claude Code on Monday that dispatches multiple AI agents to crawl pull requests, identify bugs, verify their findings, and rank issues by severity — turning what has historically been a human bottleneck into an automated triage pipeline. As @claudeai announced, the feature is now available to all Claude Code users.
The numbers Anthropic shared about its own internal deployment are striking. According to @bcherny, an Anthropic engineer, the company's code output has increased roughly 200% since adopting AI-assisted development workflows — but PR reviews had become the critical bottleneck. Before deploying the multi-agent review system internally, only about 16% of pull requests received timely review. After deployment, that figure jumped to 54%, more than tripling throughput without adding headcount to the review process.
Get our free daily newsletter
Get this article free — plus the lead story every day — delivered to your inbox.
Want every article and the full archive? Upgrade anytime.
No spam. Unsubscribe anytime.