> ## Documentation Index
> Fetch the complete documentation index at: https://docs.opsmatic.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Performance Metrics

> Understand and interpret the key performance indicators Opsmatic tracks

## Overview

This page defines the metrics you'll see across the dashboard and the [Analytics API](/api-reference/analytics), and how to interpret them. Every metric is available at daily, weekly and monthly granularity as a pre-computed aggregation.

## Core Metrics

<AccordionGroup>
  <Accordion title="Success Rate">
    **Definition:** successful executions ÷ total executions, as a percentage.

    **Watch for:** sudden drops, or a slow decline that never quite triggers a hard failure. A workflow at 92% that used to sit at 99% is quietly losing one run in twelve.

    Metric types: `daily_success_rate`, `weekly_success_rate`, `monthly_success_rate`.
  </Accordion>

  <Accordion title="Average Duration">
    **Definition:** mean execution time over the period.

    **Watch for:** upward drift (often a growing dataset or a slow upstream API) and spikes that line up with failures.

    Metric types: `daily_avg_duration`, `weekly_avg_duration`, `monthly_avg_duration`.
  </Accordion>

  <Accordion title="Execution Volume">
    **Definition:** count of executions over the period.

    **Watch for:** volume dropping to zero (a workflow that stopped triggering) and unexpected surges (a runaway trigger or retry storm).

    Metric types: `daily_executions`, `weekly_executions`, `monthly_executions`.
  </Accordion>

  <Accordion title="Availability">
    **Definition:** share of health checks where the connected platform responded successfully.

    Metric types: `daily_availability`, `weekly_availability`, `monthly_availability`. See [Availability Monitoring](/analytics/availability-monitoring).
  </Accordion>

  <Accordion title="OpenClaw Cost">
    **Definition:** rolled-up AI token spend for OpenClaw connections.

    Metric type: `monthly_openclaw_cost`. See [OpenClaw setup](/connections/openclaw-setup).
  </Accordion>
</AccordionGroup>

## Reading Trends

<Tabs>
  <Tab title="Daily">
    Best for real-time operations and catching today's incident. Highest resolution, shortest history.
  </Tab>

  <Tab title="Weekly">
    Best for spotting patterns — day-of-week effects, gradual regressions, capacity trends.
  </Tab>

  <Tab title="Monthly">
    Best for reporting and long-horizon planning. Survives raw-data retention limits because it's pre-aggregated.
  </Tab>
</Tabs>

## Querying Metrics

Pull any metric over a time range via the API:

```bash theme={null}
curl -H "Authorization: Bearer ops_your_api_key" \
  "https://api.opsmatic.com/v1/analytics?metric_type=daily_avg_duration&connection_id=conn_abc123&since=2026-06-01T00:00:00Z"
```

Results use cursor pagination — see [Cursor Pagination](/api-reference/analytics#cursor-pagination).

## Beyond Averages: Anomalies

Averages hide short spikes. Opsmatic also runs statistical **anomaly detection** over response times, failure rates and OpenClaw spend against a rolling 7-day baseline (3σ / 5σ thresholds), opening an incident when something deviates sharply.

```bash theme={null}
curl -H "Authorization: Bearer ops_your_api_key" \
  "https://api.opsmatic.com/v1/anomalies?connection_id=conn_abc123"
```

## Related

<CardGroup cols={2}>
  <Card title="Workflow Analytics" icon="workflow" href="/analytics/workflow-analytics">
    Per-workflow execution detail
  </Card>

  <Card title="Health Monitoring" icon="heart-pulse" href="/dashboard/health-monitoring">
    Real-time status on your dashboard
  </Card>
</CardGroup>