Skip to main content
Alert rules let you get notified immediately when something goes wrong in your application — without having to check the dashboard manually. Argus supports two rule types: new issue (fires the first time a distinct error appears) and error rate (fires when events cross a threshold within a time window). When a rule matches, Argus fires every configured notification channel within seconds. Configure an email address, a webhook endpoint, or both, and Argus handles the delivery.

Alert rule fields

When you create an alert rule, you configure the following fields:
FieldTypeRequiredDescription
namestringA human-readable label for this rule, e.g. "Production errors"
typeenumNEW_ISSUE (default) fires when a new distinct issue is created. ERROR_RATE fires when event volume crosses a threshold in a time window.
thresholdnumberERROR_RATEEvent count that triggers an error-rate rule within its window.
windowMinutesnumberERROR_RATERolling time window (minutes) an error-rate rule counts events over.
notifyEmailstring✅ (or webhookUrl)Email address to send a notification to when the rule fires
webhookUrlstring✅ (or notifyEmail)URL that receives a POST request when the rule fires
enabledbooleanDefaults to true. Set to false to pause the rule without deleting it.
lastTriggeredAtdatetimeTimestamp of the most recent time this rule fired a notification. Read-only.
You must supply at least one of notifyEmail or webhookUrl — a rule with neither delivery channel is rejected with a validation error. You can supply both to notify via two channels simultaneously.

Alert types

New issue (NEW_ISSUE) — fires the first time Argus sees a given error fingerprint in your project (when a brand-new issue is created). Naturally deduplicated: each distinct issue alerts once. Error rate (ERROR_RATE) — fires when the number of events in your project crosses threshold within the last windowMinutes. To avoid noise, a rule that has fired won’t fire again until one full window has elapsed (a cooldown), so a sustained spike alerts once rather than on every event.

Creating an alert

1

Open the Alerts page

Navigate to your project in the Argus dashboard and click Alerts in the sidebar.
2

Click New Rule

Click the New rule button in the top-right corner of the Alerts page. The alert rule modal opens.
3

Name your rule and choose a type

Enter a descriptive name, then pick a type: New issue to fire on every brand-new issue, or Error rate to fire on a spike.
4

Set a threshold (error-rate rules only)

If you chose Error rate, set the threshold (event count) and window (minutes) — for example, 100 events in 5 minutes.
5

Add a delivery channel

Enter an email address, a webhook URL, or both. Argus requires at least one.
6

Save the rule

Click Save. The rule is active immediately — the next new issue to arrive will trigger a notification.

Webhook payload

When a rule fires, Argus sends an HTTP POST with a JSON body to your webhookUrl. The type field tells you which kind of alert fired. Use this to integrate with Slack (via incoming webhooks), PagerDuty, or any custom pipeline you operate. New issue:
Error rate:
Your endpoint should respond with a 2xx status code to acknowledge receipt. Argus records the outcome of every delivery attempt in the alert log (see Alert logs below).

Email notifications

When a rule fires an email, Argus delivers a notification to the configured address. New-issue emails include the issue title, level, culprit, and a direct link to the issue; error-rate emails include the event count, threshold, and window, with a link to the project’s issues. Make sure the email address you configure is deliverable — Argus does not verify addresses at rule creation time.

Managing rules

From the Alerts page you can:
  • Edit a rule — click the pencil icon on any rule card to update its name, delivery channel, or enabled state.
  • Toggle enabled — click the toggle switch on a rule card to pause or resume it without deleting it. Paused rules appear dimmed.
  • Delete a rule — click the trash icon to permanently remove the rule and all of its logs.
You can also manage rules programmatically via the REST API:

Alert logs

Every delivery attempt — whether successful or failed — is recorded in the alert log. Each log entry captures:
  • Which alert rule fired
  • Which channel was used (email or webhook)
  • The target address or URL
  • Whether delivery succeeded
  • Any error message if delivery failed
  • The timestamp the alert was triggered
View logs directly in the dashboard by opening any alert rule. Use logs to diagnose failed webhook deliveries or to confirm that a notification was sent for a specific issue.
Use multiple alert rules to route notifications by severity. For example, create one rule that sends a webhook to your on-call channel for all new issues, and a separate rule with a different email address to notify a specific team member. Combine this with issue-level filters in your downstream tooling to triage by priority.
For the full list of alert endpoints, see the Alerts API reference.