quality-run

Automated repository quality scans detect dead code, security exposures, and performance smells by reading volon.yaml configurations. The scan executes enabled modes like dead_code, security, correctness, and perf_smells to identify broken links, stale tasks, sensitive files, and unreferenced skills.

Yes, you can create backlog tasks from code quality findings by configuring the on_issue action to create_task in volon.yaml. The scan writes prioritized security and correctness findings as task files to the configured storage, enabling engineering review without modifying source files.

No, the quality scan does not auto-fix security findings. It performs read-only analysis of the repository and explicitly avoids auto-fixing security issues, only writing task artifacts to the configured storage while maintaining read-only invariants.

The best way to configure multi-mode scanning for stale tasks and broken links is through volon.yaml. This configuration file determines enabled modes like dead_code, correctness, and perf_smells, while also defining per-finding actions such as log, create_task, or auto_fix_pr with configured fallbacks.

Scheduled quality audits prevent duplicate tasks by checking configured storage before writing new task artifacts. The scan enforces read-only invariants, applies configured on_issue actions, and ensures only new findings are converted into backlog tasks for engineering review.

Automated repo quality scans for sensitive files are limited to read-only detection and never modify source files. They avoid auto-fixing security findings, and while they can log issues or create task files, all source code modifications require manual engineering intervention.