Alerts

Create alert rules that fire when your logs match specific conditions. When a rule fires it opens an issue, and that issue is delivered through your project’s notification channels — Slack, Discord, email, or browser push. Delivery is configured once under Settings → Notifications and shared by every rule (and by ScryWatch’s automatic error/crash detection), so you set up your channels a single time rather than per alert.

What you’ll need

• At least one project with events flowing in
• A notification channel configured under Settings → Notifications (Slack or Discord webhook, an email address, or browser push)

Step 1: Configure where alerts are delivered

Before creating rules, set up delivery once: go to Settings → Notifications and connect the channels you want — a Slack or Discord webhook, notification emails, and/or browser push. You can also set the minimum severity and a cooldown there. Every alert rule and every auto-detected issue uses these channels.

Step 2: Navigate to Alerts

Click Alerts in the left sidebar. You’ll see a list of your alert rules (empty if you haven’t created any yet).

Step 3: Create an alert rule

Click New Rule in the top right. A rule is pure detection — it has no delivery settings of its own:

• Field — which log field to watch: level, type, service, environment, or device_type
• Value — the value to match, e.g. error, crash, payment-service
• Threshold — how many matching events must occur before the rule fires (e.g. 5 events)
• Window — the time window to count events in (e.g. 5 minutes)
• Cooldown — minimum time between repeated firings of this rule (prevents spam)

Step 4: Set a meaningful threshold

Avoid setting thresholds too low — a threshold of 1 error event will fire constantly on any active system.

Good starting points:

• Error spike: level=error, threshold=10, window=5 minutes
• Crash detection: type=crash, threshold=1, window=1 minute (crashes are always urgent)
• Service degradation: service=payment-api, level=error, threshold=5, window=5 minutes

Step 5: Enable and test the rule

After saving, the rule is enabled by default. Toggle it off using the Enabled/Disabled switch if you want to pause it without deleting it.

To test: trigger enough matching events (e.g. send 10 error-level events in 5 minutes via the ingest API) and watch for the notification in your configured channel.

Note: The cron that checks alert conditions runs every 5 minutes. There may be up to 5 minutes of delay between events arriving and a rule firing. Issues are then delivered subject to your Notifications severity and cooldown settings.

Step 6: View alert history

The Alert History section shows a log of every time a rule fired — including the matching event count and timestamp.

You’re done

You now know how to:

• Configure your notification channels once under Settings → Notifications
• Create a detection rule with a condition, threshold, window, and cooldown
• Set appropriate thresholds to avoid alert fatigue
• View the alert firing history

Related docs

Full alert API reference — create, list, update, and delete alert rules programmatically. Notification channels are managed via the project notification settings API.