Alert Rule Generator
Builds valid Prometheus alerting rule YAML from an expression, threshold, and duration, with the annotations and labels an on-call engineer needs. Defaults push toward symptom-based alerting with a sensible hold duration instead of instantaneous threshold crossings.
Alert definition
Groups are evaluated in order and share an evaluation interval.
CamelCase, no spaces. This is what appears in Alertmanager.
Alert on the symptom a user would notice, not on a cause you happen to be able to measure.
How long the expression must stay true before it fires. This is your flapping control.
Whatever your Alertmanager routes on — commonly critical, warning, or info.
One line. This becomes the notification title.
Use templating so the notification names the affected instance and the current value.
The first thing anyone does when paged is look for the procedure.
One key=value per line. These drive Alertmanager routing.
What it does
- Produces syntactically valid Prometheus rule-group YAML
- Includes summary, description, and runbook annotations by default
- Warns on patterns that commonly cause alert flapping
What it does not do
- Does not validate that your PromQL expression returns what you expect
- Does not query your Prometheus instance
Newsletter
Inside The AI Stack Brief
A practical weekly briefing on AI engineering, infrastructure, production operations, and the technologies powering the AI stack.
One email a week. No sponsorship placements inside the technical sections. Unsubscribe in one click.