code-review-assistant

To perform a structured code review on a pull request, attach your code diff to the skill. It analyzes correctness, security, and maintainability, returning inline comments with severity levels alongside a summary report.

Checking code changes for security and maintainability issues is best handled by analyzing code logic against stated intent and edge cases. This approach identifies vulnerabilities and maintainability flaws in a strict priority order.

Yes, you can use this for pre-merge checks and second opinions on code modifications. It evaluates code modifications by analyzing logic against potential edge cases to catch bugs before integration.

Automated code review handles error handling and correctness by prioritizing analysis of code logic against stated intent. It identifies critical issues beyond basic linting and categorizes them by severity levels like Blocking or Should-fix.

Severity levels assigned to issues found during code review include Blocking, Should-fix, Nit, and Question. These prioritized categories help developers understand the urgency of correctness, security, and maintainability feedback.

A limitation of automated code review for debugging is that it requires the attached code diff and stated intent to function. It focuses on logic and edge cases rather than replacing runtime debugging or basic linting checks.