OrcaCode Review
An automated code-review tool that reviews every pull request with a model you choose, posts findings inline on the affected lines, and blocks a merge when it finds a serious issue.
OrcaCode Review runs on top of OrcaRouter, reviewing every push, posting findings on the affected lines, and using a merge gate that blocks serious (P0/P1) issues while letting lower-severity findings through as comments.
A single install command detects which coding agents are present, installs matching skills for them, and configures the review from there through plain-language requests, such as asking it to review changes locally, explain why a review did not run, retune the merge policy to block only the most serious issues, or remove itself from the repository. It supports 36 agent platforms, including Claude Code, Cursor, Codex, OpenCode and GitHub Copilot, detected automatically, with the same choices available as command-line flags for CI or dotfiles use.
It installs through npx or as a Claude Code plugin, and is MIT licensed.