review-rules

Automated code review for pull requests inspects changed files for correctness, security, performance, and style. It classifies findings by severity—blocker, warning, or suggestion—so reviewers can focus on critical issues before merge.

Severity-based code review categorizes issues into blocker, warning, and suggestion ranks. This prioritization helps reviewers focus on critical correctness, security, and performance problems first, reducing regressions and avoidable rework before merge.

Post-build quality checks validate test coverage and resilience by inspecting changed code across frontend and backend paths. The review applies multi-lens inspection to report real issues, ensuring code meets quality standards after building.

Automated code review works across frontend, backend, and shared code paths. It applies multi-lens inspection to check correctness, security, performance, and style, while requiring visual verification specifically for UI changes.

You can check UI changes during pull request review by requiring visual verification. The review process inspects style and visual correctness alongside security and performance, ensuring frontend changes are validated before merge.

The best way to prevent regressions and security risks before merge is standardized code review. It inspects changed files for correctness, security, performance, and style, providing severity-ranked findings to catch important issues early.