code-review

To check a pull request diff for security risks and architectural boundary violations, review the code changes using evidence-backed reasoning and scenario-based checks to identify material defects and enforce merge policies.

Detecting hallucinated APIs and dependency risks in generated code requires verifying the existence of APIs, configurations, and libraries, grounding every review decision in explicit file evidence or named failure modes.

Enforce test coverage and error handling validation for refactored behavior by ensuring material changes have adequate tests and that failures are handled explicitly without misleading success or swallowed errors.

Yes, this code review approach classifies defect severity for production changes by categorizing findings as Critical, High, Medium, or Low and enforcing blocking rules for security and user-impacting failures.

When upgrading dependencies or modifying production code, review security surfaces including authentication, authorization, injection vectors, data exposure, crypto misuse, and error stack trace handling, documenting non-findings when applicable.

Automated evidence-grounded reasoning for code reviews is not suitable when you lack explicit merge policy expectations or when evaluating changes that do not involve material behavior, security surfaces, or architectural boundaries.