Code Review

**Judgement:** High-quality, mature, and broadly applicable code review framework. It offers a very strong, security-forward checklist and a disciplined review process that should noticeably improve review consistency and depth for most teams. **Key strengths / risks** - **Strengths:** - Excellent coverage of security, performance, and correctness concerns that are common in typical service / web stacks. - Clear three-pass process plus severity tags makes it easy to distinguish blocking from non-blocking issues and avoid random scanning. - Concrete examples of good comments and review anti-patterns help improve reviewer behavior, not just technical coverage. - **Risks / Limitations:** - Generic, language- and framework-agnostic: it won’t catch highly domain-specific or stack-specific issues (e.g., GPU kernels, embedded systems, specialized data pipelines) without additional guidance. - Checklist is dense; for very small or trivial PRs it can feel heavy and may encourage over-review if applied rigidly. - Relies on the reviewer (or model) to actually follow the structure; if ignored, you lose most of the benefits. **Recommended scenarios** - Teams formalizing or upgrading their code review standards and wanting a security- and reliability-focused baseline. - Regular review of non-trivial PRs in backend/web services, APIs, and user-facing applications where security, correctness, and performance matter. - Training and mentoring less-experienced reviewers to move from superficial to systematic reviews. - As a reusable checklist/prompt when you want the model to perform a thorough PR review with explicit severity-tagged comments.