Telemetry plane

Operational metrics for governed capability invocations and agent posture.

In buildCatalogue status

ObservingRuntime posture

Tenant-scoped capability invocation telemetry and computed agent SLO summaries.

Open console Book a Demo

Expanded capabilities

Expose operational telemetry and read-model health for the tenant
Keep freshness/staleness explicit on command surfaces
Support HERMES/TERMINUS observation without mutating jobs
Avoid turning missing metrics into decorative zeros
Feed SLO-minded operator attention without silent autoscaling claims

How they learn & improve

Telemetry improves which gauges matter from operator pins — it does not auto-remediate production.

Collect tenant-scoped operational counters with freshness.
Surface stale/missing gauges honestly.
Recalibrate attention after human-applied learning on noisy alerts.

How they work together

Telemetry feeds HERMES flow health and TERMINUS substrate views so operators correlate signals instead of chasing separate dashboards.

HERMESConsumes telemetry into flow-health narratives.
TERMINUSPairs counters with ExecutionJob substrate posture.
Operator console (Next.js)Renders gauges on operator surfaces.
AUDITKeeps meaningful ops actions attributable.

Practical benefits

Stale is labelled staleOperators don’t trust yesterday’s counter as ‘live green’.
Correlated ops storyFlow + substrate + gauges sit together — fewer tool-hops during incidents.
Quiet the noise deliberatelyLearning candidates can demote noisy gauges after human apply — not by agents muting themselves.

Learning and improvement stay human-visible: candidates are RECORD_ONLY until an operator reviews and applies them. Sensitive sends, publishes, and approvals never run silently.

Telemetry plane

Expanded capabilities

How they learn & improve

How they work together

Practical benefits

Related agents