OrcaCode Review — the open code review harness. Multi-model reviews, merge gates, and no markup. Pay only for inference.
Rules and checks that make Claude Code look around a change, not just at the lines it writes: ten questions before code, a reviewer that didn't write it, proof before done, bugs fixed as a class.
20-pass adversarial audit for vibe-coded software. Injection, IDOR, races, N+1, idempotency, transactions, memory leaks — then a false-positive verifier.
A Stop hook that won't let the turn end until fresh-context reviewers have read the diff, their findings are verified, and the loop converges — or honestly says it didn't.
I review of Lustre patches in VS Code: local commits and Gerrit changes, findings in the diff, publishing to Gerrit
AI-powered GitHub App that automatically reviews every pull request - catching security vulnerabilities, bug risks, and architectural issues as inline PR comments. Node.js · OpenAI gpt-4o-mini · React · MongoDB
A centralized code-review platform for GitHub built on Claude. One shared engine powers three agents - Security (auto on every PR), Code Review (auto, quality-only, toggleable), and Branch Review (manual, full branch). Target repos add a tiny caller workflow and inherit everything, fix the logic once and all repos benefit.
To associate your repository with the code-review-assistant topic, visit your repo's landing page and select "manage topics."