Platform

Platform overviewWorkflow orchestrationGitOps configurationGovernance and AAAAI and MCPRuntime 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
Pricing

Company

How a run worksAboutContactBook a demo

06 · Evidence and monitoring

Keep operational evidence attached to the work.

NopsAI stores durable pipeline evidence and exposes platform health through bounded logs and Prometheus metrics. Monitoring views relate execution behavior to the workflow, runner, trigger, schedule, model and output.

Run ownedLogs, history, approvals, context and outputs
Platform wideSystem logs, health, runners and metrics
ActionableViews, alerts, recommendations and audit

How it works

Evidence and monitoring, end to end.

The flow stays explicit enough for operators to inspect, secure, automate and improve.

  1. Capture run evidence

    Persist normalized trigger data, authorization context, task and step transitions, logs, approvals, knowledge snapshots and outputs.

  2. Collect platform signals

    Read allow-listed Docker or Kubernetes service logs and expose Prometheus metrics from existing service owners.

  3. Analyze operating behavior

    Aggregate run, pipeline, step, task, trigger, schedule, model, runner, reliability, efficiency and security views.

  4. Drive follow-up

    Save monitoring views, evaluate alert rules and track recommendations through acknowledgement and resolution.

Pipeline runs overview with running-now, needs-attention, success rate and median duration cards above a live run list.

Platform capabilities

What teams can standardize.

These are current NopsAI platform primitives, not a future-state feature list.

Durable execution history

Inspect run status, task history, logs, child runs, approvals, resolved context and deliverables together.

System logs

Stream bounded, redacted service logs through authenticated SSE with reconnect and cursor behavior.

Operational analytics

Compare pipeline, step, task, trigger, schedule, runner, reliability, efficiency and security behavior.

AI and output audit

Track model usage plus final-output generation, contract correction, rendering and persisted file state.

Configuration

Declared once, reviewed like code.

The example is declarative and Git-reviewable. Runtime behavior still passes through live AAA, credential, compatibility, execution and monitoring boundaries.

config-repositories/teams/platform/notifications.yamlYAML
enabled: true
routes:
  - name: production failures
    enabled: true
    recipients:
      include:
        teams:
          - platform/prod
        users:
          - [email protected]
    events:
      failure: true
      approval_requested: true
      approval_rejected: true
      cancelled: true
    filters:
      pipelines:
        include:
          - platform/release/*
      branches:
        include:
          - main
          - release/*
    delivery:
      channels:
        - mail
      throttle:
        dedupe_window: 10m
        max_per_run: 5

Enterprise controls

Configuration stays inside the governance model.

Content-free stream audit

Access to system log streams is audited without copying streamed log content into the audit record.

Least-privilege providers

Docker uses a restricted socket proxy and Kubernetes uses label-selected, read-only pod log permissions.

Redaction and limits

Log sources, line size, buffers, cursors and sensitive patterns are bounded by the system-log owner.

GitOps-aware operations

Monitoring definitions and notification routes can be represented as reviewable platform configuration.

Bring a real evidence and monitoring workflow.

We can map the configuration, authorization, runtime and evidence boundaries against your current platform model.