system_base_manager

System health auditing detects configuration drift, broken references, and sync gaps across agent infrastructure by scanning directories like ~/.claude/ and ~/agents/agent-system-base/ to identify inconsistencies and output recommended corrective actions.

Auditing stalled runs and infrastructure gaps in an orchestrator system involves scanning agents, skills, plans, schemas, runs, and scripts to detect stale processes and trigger proactive audits for full lifecycle management.

Running a complete system audit across agents, skills, plans, orchestrator, dashboard, schemas, runs, and scripts scans the infrastructure directories to detect broken references, sync drift, and stale runs, then outputs audit findings with recommended actions.

Broken references and stale runs after updating skills occur due to sync drift across the agent infrastructure; auditing detects these inconsistencies by validating reference integrity across agents, skills, plans, orchestrator, schemas, runs, and scripts.

Yes, full system lifecycle management tracks audit history and maintains health across all components including agents, skills, plans, orchestrator, dashboard, schemas, runs, and scripts, extending beyond skills to the entire infrastructure.