code-review

Perform a structured code review by evaluating pull request changes for correctness, security, quality, testing, and maintainability. The process outputs classified findings like BLOCKER or SUGGESTION tied to file and line, ending with an APPROVE or REQUEST CHANGES decision.

A code review checklist ensures consistent evaluation before merging by covering correctness, security risks like secrets and injection, readability, duplication, complexity, and test presence. It classifies findings into BLOCKER, WARNING, SUGGESTION, or PRAISE to guide the final merge decision.

Identify security and maintainability issues in a diff by checking for hardcoded secrets, missing input validation, injection vulnerabilities, sensitive logs, high complexity, and code duplication. Findings are categorized by severity and pinpointed to specific file lines for actionable remediation.

Yes, the code review process applies to pull requests, branches, and diff-scoped edits across multiple languages and frameworks. It requires an explicit review context to understand the diff scope and intent before assessing correctness, security, quality, and maintainability.

Request changes on a pull request when the review identifies any BLOCKER-level findings. The final decision mandates REQUEST CHANGES if blockers exist, while WARNING or SUGGESTION findings may result in a COMMENT or APPROVE if no blocking issues are found.

Automate code review for branches and diff-scoped edits by applying a structured checklist that evaluates context, intent, and code quality. The automated review generates categorized findings and a final merge decision based on the presence of blocking issues.