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

# Monitoring Rules

> Define what signals should trigger alerts, for which scope, and through which channels

Monitoring rules define when PG:AI should generate alerts and who should receive them.

## Rule components

* **Source type**: what to monitor (jobs, employee changes, custom columns, external events, etc.)
* **Condition type**: how to evaluate change (threshold, equals, created, changed, etc.)
* **Scope**: organisation-wide, territory, or selected companies
* **Frequency**: real-time, hourly, daily, weekly, or on-change
* **Channels**: in-app, email, Slack, webhook

## Recommended rollout

1. Start with a small set of high-value rules.
2. Set severity and cooldowns to reduce noise.
3. Review trigger volume after one week.
4. Tighten conditions before adding more rules.

<Warning>
  Over-broad rules create alert fatigue quickly. Start narrow, then expand once signal quality is proven.
</Warning>