Uptimerobot

@uptimerobot show me all monitors that are currently down or paused, including their response times over the last 24 hours

Based on your Uptime Robot account, 2 monitors are currently flagged: 'API Gateway (prod)' is down (last checked 3 minutes ago, HTTP 503 response), and 'Payment Processor' is paused for maintenance. Over the last 24 hours, the API Gateway averaged 1,240ms response time before going down; the payment processor was paused 6 hours ago so no recent data is available. All other 12 monitors are up with average response times under 400ms.

This example demonstrates read-heavy querying across multiple monitors. Requires an API key with read access to all monitors. Response time granularity depends on your Uptime Robot plan tier—free accounts get 5-minute checks, paid plans get 1-minute intervals. The MCP fetches current state; historical trend analysis beyond 24 hours may require additional calls.

@uptimerobot create a new HTTPS monitor for https://checkout.example.com checking every 5 minutes, and alert me via email if it goes down for more than 2 consecutive checks

Monitor created successfully. 'checkout.example.com HTTPS Monitor' (ID: 794382156) is now active, checking https://checkout.example.com every 5 minutes via HTTPS. Alert contacts configured: your account's default email will receive notifications after 2 consecutive down checks (approximately 10 minutes of downtime). The monitor is live and will perform its first check within the next 5 minutes. You can view real-time status in your Uptime Robot dashboard.

This example shows the MCP's ability to provision new monitoring with specific intervals and alert thresholds. Requires an API key with write permissions. Check interval options depend on your plan—free accounts are limited to 5-minute intervals. The MCP creates the monitor but doesn't validate the target URL is reachable; you'll see the first check result after the interval elapses.

@uptimerobot review all monitors tagged 'production' and draft a status page summary showing which services are healthy, which have had incidents this week, and overall uptime percentage

Analyzing 8 production-tagged monitors: 6 are currently healthy with 99.8%+ uptime this week. Two incidents detected: 'Auth Service' experienced a 14-minute outage on Tuesday (99.86% weekly uptime), and 'CDN Edge Node' had three brief downtimes totaling 8 minutes (99.92% weekly uptime). Suggested status page summary: 'Production systems operating normally. Auth Service and CDN experienced brief interruptions earlier this week, now resolved. Overall production uptime: 99.89% (7-day average).' Consider scheduling maintenance windows for the Auth Service to investigate recurring timeout patterns.

This synthesis example pairs Uptime Robot data with AI reasoning to generate stakeholder-ready summaries. The MCP fetches raw monitor states and logs; the AI interprets patterns and drafts prose. Accuracy depends on how monitors are tagged in your account—ensure consistent tagging conventions. Historical incident data is limited by your plan's log retention (typically 2 months for paid plans).

When Uptimerobot beats building your own status page

A 10-person SaaS startup needs a public status page for their three core services (API, web app, background jobs). Uptimerobot is the right call here. You get monitor creation, public status page editing, and maintenance window scheduling in one MCP—no custom dashboard to build or host. The 18 tools cover the full lifecycle: add monitors for each service, configure a branded status page, schedule maintenance windows when you deploy. The trade-off: if you need sub-minute granularity or custom incident workflows, you'll outgrow this fast. But for a team that wants a working status page this week without hiring a DevOps engineer, this MCP delivers.

Uptimerobot for weekly monitor audits during handoff

A 6-person agency runs client sites on shared infrastructure. Every Monday, the on-call engineer rotates and needs to verify all 40 client monitors are active and correctly configured. The Get Account Details and Edit Monitor tools let you script this handoff: pull the full monitor list, flag any with stale URLs or disabled alerts, update contact lists when clients churn. The MCP works because the agency's monitoring needs are stable—same check types, predictable cadence. The boundary: if you're adding 10 new clients a month, the manual monitor setup (even via Add Monitor) gets tedious. At that scale, you want Terraform or a bulk import API. For a stable client roster, this MCP turns handoff into a 10-minute Switchy workflow.

When support needs read-only monitor context, not full control

A 4-person support team fields downtime tickets but doesn't own the monitoring stack. They need to check if an outage is real or user error—did the monitor actually fire, or is the customer's network flaky? The Get Maintenance Windows and Get Monitor Authentication Type tools give support read access to confirm scheduled maintenance or auth issues without touching live configs. The MCP fits because it's scoped to lookup, not mutation. The catch: if your support team needs to acknowledge incidents or update status pages directly, Uptimerobot's MCP doesn't expose those workflows—you'd need PagerDuty or Opsgenie integration instead. For pure triage context, this MCP keeps support informed without handing them the keys to production monitoring.