> ## 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.

# Health Monitoring

> The real-time monitoring widget and how Opsmatic tracks platform health

## Overview

Health monitoring is Opsmatic's real-time pulse on your automation platforms. The dashboard's monitoring widget shows current status and the most recent checks, and the monitoring pipeline opens incidents and sends alerts when something breaks.

## The Monitoring Widget

The widget on your main dashboard surfaces, at a glance:

<CardGroup cols={2}>
  <Card title="Current status" icon="signal">
    Whether each connection is up right now.
  </Card>

  <Card title="Latest checks" icon="list-check">
    The most recent availability results and response times.
  </Card>

  <Card title="Open incidents" icon="triangle-exclamation">
    Anything currently degraded, with a link to the incident.
  </Card>

  <Card title="Heartbeats" icon="wave-square">
    For OpenClaw, the last heartbeat received from each gateway.
  </Card>
</CardGroup>

## How Health Is Determined

<Steps>
  <Step title="Scheduled checks">
    Each connection is checked on a tier-based interval. Availability probes confirm the platform responds; OpenClaw gateways report in via heartbeats.
  </Step>

  <Step title="Failure detection">
    A run of failed checks — or a missed heartbeat — opens an incident of the relevant kind (`availability`, `missed_heartbeat`, `anomaly`, `budget`).
  </Step>

  <Step title="Alerting">
    Failure and recovery emails go out, with a link to the incident and its timeline.
  </Step>

  <Step title="Auto-resolve">
    Availability incidents close automatically after two consecutive successful checks.
  </Step>
</Steps>

## Incident Kinds

| Kind               | Trigger                                     |
| ------------------ | ------------------------------------------- |
| `availability`     | Platform failed its checks                  |
| `missed_heartbeat` | An OpenClaw gateway stopped reporting       |
| `anomaly`          | A metric deviated sharply from its baseline |
| `budget`           | An OpenClaw spend threshold was crossed     |

Each incident moves through `open → acknowledged → resolved` and keeps an append-only event timeline.

## Checking Health via API

```bash theme={null}
# Current open incidents
curl -H "Authorization: Bearer ops_your_api_key" \
  "https://api.opsmatic.com/v1/incidents"

# Latest availability samples for a connection
curl -H "Authorization: Bearer ops_your_api_key" \
  "https://api.opsmatic.com/v1/availability?connection_id=conn_abc123"
```

## Related

<CardGroup cols={2}>
  <Card title="Availability Monitoring" icon="server" href="/analytics/availability-monitoring">
    Uptime tracking in depth
  </Card>

  <Card title="Connection Problems" icon="wrench" href="/troubleshooting/connection-problems">
    What to do when a connection is unhealthy
  </Card>
</CardGroup>