MCPHub Docs

Alerts

Configure alert rules that fire when servers go down, error rates spike, or latency climbs.

Alerts

MCPHub's alert system notifies you when something goes wrong with your MCP servers. You define rules with conditions and thresholds; MCPHub evaluates them against recent data and fires alerts when conditions are met.

Alert lifecycle

Alert rule created (active)

Alert evaluator runs

Condition met → Alert fires (FIRING)
  Notification sent (Slack / webhook)

Condition clears → Alert resolves (RESOLVED)
  Resolution notification sent

When alerts are evaluated

Alert evaluation runs:

  • Daily at 02:00 UTC via Vercel Cron
  • On demand — click Evaluate Alerts on the Alerts page (admin/owner only)

In this section

On this page