Claude Code Review Goes GA With Per-Repository Review Depth
Anthropic moved Code Review out of research preview on October 15 and made it generally available to Claude Team and Enterprise organizations. The change that matters is that review depth is now a per-repository setting: admins can set each repository to Quick, Standard (the default) or Deep, so a routine internal tool gets a light pass while a payments service gets a thorough one. Anthropic’s announcement has the details.
Anthropic says most reviews now finish about three times faster, down from roughly 20 minutes to 6-10 minutes, and that findings state the problem, what triggers it and the fix in a few sentences. Those are the vendor’s own figures from preview teams, not an independent benchmark, so test them on your own repositories before you plan around them.
The reason this matters is volume. Anthropic’s engineers wrote in September that they now ship 8x as much code per quarter as they did from 2021 to 2025, with Claude authoring 80% of it, and that CI jobs surged 25x in six months. When code arrives that fast, human review becomes the next bottleneck after the test suite. A reviewer that runs automatically on every pull request is infrastructure, not a convenience.
The decision for a leadership team is not whether to switch it on. It is which repositories earn Deep review and who owns the findings. Setup is an admin task: enable Code Review in Claude Code settings, install the GitHub App and select repositories. Developers configure nothing. Eligible Claude Enterprise organizations can claim at least $10,000 in Code Review credits by January 15, 2027, and credits expire 30 days after claiming.
Start with the two repositories where a missed bug costs the most, set them to Deep, and measure what the reviewer catches that people did not. That is how you find out whether it ships outcomes or just adds comments.