requesting-code-review

Pre-commit code review on staged git diffs combines static security scanning for hardcoded secrets and injection risks with baseline-aware lint checks. An independent reviewer evaluates the changes and returns a strict pass or fail verdict before committing.

Baseline-aware test gating runs relevant test suites and blocks commits only for new regressions introduced by your staged changes. It counts pre-existing baseline failures separately, ensuring your diff is evaluated solely on the new issues it creates.

Yes, a bounded auto-fix loop patches reported security and lint issues by delegating to a separate fix agent. After patching, the verification pipeline re-runs static scans, tests, and lint checks to confirm the fixes resolved the original problems.

A fail-closed independent reviewer is a separate subagent that analyzes your staged git diff and outputs a strict JSON verdict. If the verdict indicates issues or the format is invalid, the pipeline blocks the commit to prevent unverified code from landing.

Automated pre-commit verification is ideal for git-based workflows after feature or bugfix tasks involving multiple file edits. It is especially useful following subagent-driven development steps to ensure code quality, security, and logic are validated before changes land.

Pre-commit security scanning detects common red flags like hardcoded secrets, shell injection patterns, eval/exec usage, unsafe deserialization, and SQL injection risks. It specifically targets the added diff to prevent these vulnerabilities from being committed.