fix-review-code

To automatically fix code review issues, you can run the fix-review-code command on your branch with the target commit. It reads an existing review report or generates one to correct naming, style, and security violations with safety guardrails.

Yes, you can preview code refactoring changes by running the command with the --dry-run flag. This non-modifying mode shows the intended fixes for type safety and pattern violations without altering your files or creating scoped commits.

The best way to prioritize security and type safety fixes is using severity-ordered remediation. The tool processes issues from CRITIQUE to BASSE, ensuring the highest-impact violations are corrected first before generating a structured fix summary.

Yes, the automated fixing tool requires interactive confirmation by default to ensure safety. You can bypass this for non-ambiguous fixes by using the --auto flag, which applies strategy-driven edits directly.

To handle ambiguous code violations during refactoring, the tool enforces strict ambiguity handling modes and maps each issue to deterministic strategies. This ensures imports and pattern edits are resolved safely without guessing.

A key limitation is that auto-fixing requires matching the exact commit and branch parameters used in the original code review. Additionally, ambiguous issues cannot be auto-fixed and require interactive confirmation to proceed safely.