Diagnostics
The Diagnostics page shows operational signals from the running server. Only system administrators can open it. The signals are read-only. Warning thresholds can be adjusted without changing hard safety limits.
The page groups signals into tabs:
| Tab | Shows |
|---|---|
| Server clock | The offset between the server clock and the browser clock. A rolling median of the latest samples corrects displayed timestamps and warns when the offset passes the threshold. |
| Action executions | Recent automation action runs and their failures. |
| Webhook deliveries | Recent outbound webhook deliveries. |
| Background jobs | The scheduler jobs and their last runs. |
| Recurrence | Recurring item rules and their next runs. |
| Domain events | The internal domain event backlog. |
| Frac index | The fractional index used for manual ordering. |
| Runner pools | Registered coding-agent runners and their capacity. |
| Database pools | Connection pool use for the database. |
| Cache memory | In-process cache use. |
| SCM connections | Source control connections and their state. |
Use the page when work is delayed, a webhook is not delivered, or a scheduled job does not run. The server clock tab explains a timestamp difference between the server and a user's browser.
Related documentation
- Actions for action configuration.
- Webhook channels for outbound webhooks.
- Prometheus metrics for external monitoring.