olOpsLyft Docs

Detection rules

Control what counts as an anomaly and what happens when one is found.

Detection rules decide which changes in spend become anomalies. OpsLyft ships with default rules that cover every connected provider.

Manage rules

On Monitor → Anomalies, open + Detection rule:

  • New rule creates a rule.
  • Manage rules lists your rules with a count.

What a rule controls

SettingControls
ScopeProviders, accounts, services, or tags the rule watches
SensitivityHow far spend must leave its expected range to count
Minimum impactThe smallest excess per day that raises an anomaly, so small services don't page anyone
SeverityHow impact maps to High, Medium, and Low
NotifyWho hears about it: people, Slack channels, or email

Turn anomalies into tickets

Open tickets automatically with auto-create rules in Tickets → Integrations, for example:

  • High-severity anomalies → ticket — impact ≥ $200/day, to the team's project, priority High, SLA 1 day.
  • Medium anomalies → ticket — impact ≥ $50/day, to the team's project, priority Medium.

See Auto-create rules.

On this page