How Monitoring Works

Understand how we check your endpoints, detect incidents, and keep your status page up to date.

Endpoint checks

We send HTTP requests to your endpoints at regular intervals and record:

• HTTP status code — 2xx = success, 4xx/5xx = failure
• Response time — milliseconds
• Timestamp

Check intervals depend on your plan:

• Free — every 5 minutes
• Pro and Team — every minute

All checks are performed from a single region (EU). Multi-region checks are on the roadmap.

Incident detection

An incident is created automatically when an endpoint fails 3 consecutive checks. This prevents false positives from transient network issues.

When an incident is created:

• The endpoint status changes to “down”
• You receive an email alert
• If configured, a Slack notification is sent
• Status page subscribers are notified (Pro and Team)
• The incident appears on your public status page

Incident lifecycle

Incidents progress through four statuses:

1. Investigating — initial state when the incident is created
2. Identified — root cause has been found
3. Monitoring — a fix has been deployed, watching for recovery
4. Resolved — the issue is fixed

You can update the status and post timeline messages from the dashboard, MCP, or API. Each update is timestamped and shown on the public status page.

Auto-resolution

When a previously failing endpoint passes 3 consecutive checks, the incident is resolved automatically. You'll receive a resolution alert.

Flap detection

If a service creates 5 or more incidents within 1 hour, we pause monitoring for that endpoint and notify you. This prevents alert fatigue from unstable services and protects your status page from noise.

AI post-mortems

When an incident is resolved, you can generate a structured post-mortem with one click. The AI analyzes:

• Check history (response times, status codes, failure patterns)
• Timeline updates you posted during the incident
• Duration and impact scope

The generated report includes: summary, timeline, root cause analysis, impact assessment, resolution details, and prioritized action items.

On Pro and Team plans, you can enable auto-generation in project settings — the post-mortem is created automatically when an incident resolves.

Data retention

• Free — 7 days of check history
• Pro and Team — 90 days of check history

The public status page shows up to 90 days of uptime data.

SLA monitoring (Team plan)

Set a monthly uptime target (e.g., 99.9%). We track the rolling 30-day uptime for each endpoint and alert you when any endpoint drops below your target.