gsd-audit-fix

To automatically fix recurring code audit findings, you can run an audit-to-fix pipeline that classifies issues into auto-fixable and manual-only categories. It applies deterministic fixes to eligible issues based on configured severity limits and validates them with test verification before committing.

Automating CI remediation involves running an audit that targets recurring lint or security findings, applying deterministic fixes to the safe subset of issues, and leaving the rest for human review. It uses atomic commit boundaries to ensure changes remain small and reviewable.

You can limit automated code fixes by using the --severity and --max flags when running the audit. This restricts autonomous remediation to a specific severity level, such as medium, and caps the maximum number of findings processed to a configured limit like 5.

Yes, you can test code changes before committing during automated repository maintenance. The workflow integrates test checks after applying fixes and only creates small, reviewable atomic commits after the test run successfully validates the changes.

Yes, the automated audit fix workflow supports a --dry-run mode. This allows you to evaluate which findings the audit classifies as auto-fixable versus manual-only and preview the deterministic remediation without applying any changes to the repository.

Some code audit findings are skipped during autonomous remediation because the audit classifies them as manual-only rather than auto-fixable. This ensures that only issues which can be deterministically remediated based on the audit output and severity limits are automatically fixed.