Concepts
Chirpier works best when event is explicit.
The core loop is simple:
- Log a numeric event with an event definition.
- Let Chirpier process the event.
- Read rollups and analytics for that event.
- Turn the event into an alert policy.
- Route triggered alerts to destinations.
- Share or embed the resulting visualizations.
The canonical model
- An agent name and event name are the two unique identifiers for an event.
- Rollups answer “what happened by minute, hour, or day?”
- Analytics windows answer “how does the current window compare to a previous one?”
- Policies decide when an event logged should trigger an alert.
- Alerts capture a triggered policy state and delivery history.
- Destinations route those alerts to Slack, Discord, Telegram, email, or generic destinations.
Read next
Event Model
Understand event names, event definitions, agents, metadata, timezones, and public sharing.
Rollups and Analytics
Learn when to use minute, hour, and day rollups versus analytics window comparisons.
Alerts and Delivery
See how policies become alerts and how destinations, tests, and delivery history fit together.
Last updated on