Skill Repair

Automate debugging for failing workflows by running a diagnostic lifecycle that applies systemic triage, regression analysis, and PR-based fixes to restore skill reliability. The process parses cron-state, recent runs, and issue memory to execute and verify repairs automatically.

Diagnose degraded skills systemically by clustering root causes across multiple tracked skills or executing per-skill repairs. This approach applies preflight checks and regression hunting to identify and resolve reliability issues automatically.

Yes, you can automate pull-request generation for skill repairs. The diagnostic workflow creates PR-based changes and verification plans, integrating issue-tracking to maintain a complete repair history and audit trail for restored workflows.

Run a systemic repair by leaving the skill name blank during diagnosis. The workflow automatically parses recent runs and issue memory to select the worst failing target, execute fixes, and record results without manual input.

No, the repair workflow requires no dependencies or components to function. It operates independently to parse cron-state and recent runs, applying regression analysis and per-skill repairs across tracked skills with PR-based verification.

Automated skills fail after workflow changes due to regressions. The repair workflow addresses this by applying regression hunting and systemic triage across tracked skills to identify root causes and generate PR-based fixes.