OmniNode-ai System Status
Fast Path — Node Dispatch
Primary fast path dispatches to node_platform_diagnostics for runtime node and hook health:
onex run node_platform_diagnostics -- --dimensions=RUNTIME_NODES,HOOK_HEALTH
For a full 7-dimension platform check:
onex run node_platform_diagnostics -- --dry-run
Results are returned as ModelDiagnosticsResult JSON with per-dimension PASS/WARN/FAIL status.
Full platform health monitoring with diagnostics and status reporting.
Sub-skills
- check-agent-performance - Review agent routing performance metrics
- check-database-health - Verify PostgreSQL connectivity and schema state
- check-infrastructure - Check Docker and infrastructure services
- check-kafka-topics - Verify Kafka/Redpanda topic health and consumer groups
- check-pattern-discovery - Check pattern discovery pipeline status
- check-recent-activity - Review recent agent and system activity
- check-service-status - Check individual service health endpoints
- check-system-health - Full system health summary
- diagnose-issues - Diagnose common platform issues
- generate-status-report - Generate a comprehensive status report
Agent Diagnostics
Agent-specific monitoring and diagnostics capabilities, including alert thresholds and database-backed profiling.
Alert Thresholds
| Level | Success Rate | Unprocessed Events | P95 Duration | Error Rate |
|---|---|---|---|---|
| CRITICAL | < 70% | > 200 | > 120,000ms | > 30% |
| WARNING | < 80% | > 50 | > 60,000ms | > 20% |
| HEALTHY | >= 90% | < 50 | < 30,000ms | < 10% |
Database Tables Monitored
- agent_execution_logs -- Core execution tracking with status, duration, quality scores
- agent_routing_decisions -- Agent detection and routing intelligence with confidence scores
- hook_events -- Hook event processing status, retry counts, and processing errors
- agent_detection_failures -- Failed agent detection attempts for debugging routing issues
Recommended Workflows
Daily Monitoring:
/system-status check-system-health
/system-status diagnose-issues --time-range 24h # if issues detected
Incident Investigation:
/system-status check-system-health
/system-status diagnose-issues --time-range 1h
/system-status check-agent-performance --agent [failing-agent]
Weekly Review:
/system-status generate-status-report --time-range 7d