DocsBook a demo

Platform

Platform overviewArchitectureWorkflow orchestrationGitOps configurationGovernance and AAAAI and MCPKnowledge and contextRuntime and executionEvidence and monitoring

Use cases

All use casesProduction incidentRelease preparationHotfix to productionSecurity scan triage
Why NopsAIIntegrationsSecurity

Resources

All resourcesAI agent governanceMCP governanceMCP securitySelf-hosted platforms
PricingGitHub

Company

How a run worksAboutContactBook a demo

Monitoring and metrics

Authorization-filtered analytics across runs, triggers, runners, AI usage, reliability, efficiency, and security.

ReferenceOperatorAdministrator

Key points

  • Monitoring is authorization-filtered: you see the runs, pipelines, and resources you are allowed to see.
  • Coverage spans runs, pipelines, steps, tasks, triggers, external triggers, runners, LLM usage, reliability, efficiency, and security.
  • GET /metrics exposes Prometheus metrics including identity-provider capability and authorization grant ownership series.
  • Metrics are public by default for scraper compatibility and can require bearer auth with metrics_require_auth.
  • Alert rules can be created, evaluated on demand, and produce alert events.
  • Recommendations can be acknowledged or resolved so the same finding does not resurface indefinitely.

Examples

Pull the operational summarybash
curl -s -H "Authorization: Bearer $NOPSAI_TOKEN" "$NOPSAI_URL/v1/monitoring/summary" | jq
curl -s -H "Authorization: Bearer $NOPSAI_TOKEN" "$NOPSAI_URL/v1/monitoring/reliability" | jq

# Prometheus scrape target; set METRICS_REQUIRE_AUTH=true to require a token
curl -s "$NOPSAI_URL/metrics" | head -20
Result

The monitoring routes answer product questions — reliability, efficiency, AI usage — while /metrics is the raw scrape surface.

How it works

AI usage telemetry records hashes, revision markers, cache and session identity, provider-state IDs and support, cached input tokens, cache-write tokens where providers expose them, and bounded workspace retrieval sizes — without storing prompt bodies.

Saved monitoring views let a team keep a filter set rather than rebuilding it each time.

Implementation evidence

  • services/nopsai/monitoring_analytics_schema.go

    Analytics schema and filters.

  • services/nopsai/routes.go

    The monitoring route surface.