Threshold rules
Set a static condition on any metric or service state — CPU %, memory, disk, restarts, request errors. Straightforward to write and to reason about later.
Alerts & incidents
Fewer pages. Clearer incidents.
PrimesFlow only pages you when something genuinely matters, folds related noise into a single incident, and logs a full timeline so you can review — and prevent — the next occurrence.
Capabilities
Alerting built for signal, not noise
Anomaly detection
PrimesFlow builds a baseline per host and per service, then flags today's behavior when it diverges from the last two weeks — no manual threshold tuning.
Grouping & noise control
A single bad host produces a single incident, not a wall of pages. Related alerts fold together automatically.
On-call routing
Send alerts to email, Slack, Discord, generic webhooks, or PagerDuty and Opsgenie via webhook — route different severities to different destinations.
Timeline for every incident
Each incident keeps a chronological log of what tripped, what recovered, which alerts fired, and who acknowledged it.
Pattern detection over time
PrimesFlow surfaces incidents that share a signature across time so you can address the underlying cause, not just the latest occurrence.
Routing
Reaches the tools your team already lives in
Slack
Discord / Webhooks
PagerDuty & Opsgenie
Example incident timeline
Incident #4821 · web-03 · high-CPU saturation · resolved in 8 minutes.
- Trigger14:22:10 UTC
CPU on web-03 crossed 85% threshold (rolling 3m).
- Alert14:22:14 UTC
Alert routed to #ops on Slack and PagerDuty (severity: high).
- Note14:23:01 UTC
Anomaly detector: memory usage also diverging from baseline.
- Ack14:24:47 UTC
Acknowledged by @maya.
- Note14:27:52 UTC
AI diagnosis drafted — likely correlated with deploy marker at 14:20:38.
- Recovered14:30:11 UTC
CPU recovered below threshold; incident auto-resolved after 3m stable.
Cut through alert noise
14-day free trial. Bring your Slack, Discord, or PagerDuty webhook.