MCP server for your uptime monitoring

Connect your AI assistant to UptimeIO and ask about your monitors and incidents in plain language, without leaving your editor.

Why connect an assistant to your monitoring

When something breaks, you are usually already in your editor or terminal with an AI assistant open. Switching to a dashboard to find out what is down, when it started, and which locations saw it takes time you would rather spend on the fix.

The Model Context Protocol (MCP) is an open standard that lets AI assistants use tools from other services. UptimeIO runs a remote MCP server, so an assistant such as Claude Code or Cursor can read your monitors and incidents directly and answer questions like "which monitors are down?" or "what happened during last night's incident?".

On paid plans the assistant can also make changes for you, such as creating a monitor for a new endpoint or pausing one before planned work.

How it works

  1. Create an API key

    In the dashboard, go to Settings, then API Keys. Use a Read Only key unless you want the assistant to make changes.

  2. Add UptimeIO to your assistant

    Add the UptimeIO MCP server URL and your key to Claude Code, Cursor, or another assistant that supports remote MCP servers.

  3. Ask in plain language

    Ask which monitors are failing, how response times look this week, or what an incident's check attempts showed.

  4. Let it act, on paid plans

    With a Read & Write key on Pro or Scale, the assistant can create, pause and resume monitors, acknowledge incidents and schedule maintenance.

What the assistant can do

Tools follow the same rules as the dashboard. Your role decides what you can do, and a key limited to one project only reaches that project.

  • Read monitors: status, uptime, response time, and settings, with header values and request bodies masked.
  • Read response times: averages, percentiles and a breakdown of DNS, connect, TLS and first-byte time, within your plan's history.
  • Read incidents: updates, affected monitors, and the check attempts behind each incident, so the assistant can explain what failed and where.
  • Read status pages: names, public URLs and how many monitors each one shows.
  • Make changes on Pro and Scale: create, pause and resume monitors, acknowledge incidents, and schedule maintenance windows.
claude mcp add --transport http uptimeio https://api.uptimeio.com/mcp \
  --header "Authorization: Bearer uio_YOUR_API_KEY"
Example: add UptimeIO to Claude Code

How it fits with your alerts

The MCP server does not replace alerts. Incidents still open only after failed checks are confirmed from more than one location, and your alert channels are notified as usual. The assistant is a faster way to look into an incident once you know about it.

  • Email, Slack, Telegram, Pushover, and webhooks, on every plan (Pushover on Free needs your own app token)
  • The assistant sees the same incidents and check results as the dashboard
  • There is no SMS or phone-call alerting

Which plans include it

The MCP server is available on every plan. On Free, API keys are read-only, so the assistant can read monitors, incidents and status pages but cannot change anything. On Pro and Scale, a Read & Write key also allows changes.

Tool calls count against your organization's API limit: 10 per minute on Free, 200 on Pro and 500 on Scale. Keep your API key out of files you commit to source control.

  • MCP server docs: Setup for Claude Code and Cursor, the full tool list, and plan limits.

Frequently asked questions

What is an MCP server?

MCP, the Model Context Protocol, is an open standard that lets AI assistants use tools from other services. The UptimeIO MCP server gives your assistant tools to read and manage your monitoring.

Which assistants work with it?

Any assistant that supports remote MCP servers over HTTP with an API key header. The docs include setup steps for Claude Code and Cursor.

Is it available on the Free plan?

Yes. On Free the assistant can read your monitors, incidents and status pages. Making changes needs a Read & Write key on Pro or Scale.

Can the assistant delete monitors?

No. The assistant can create, pause and resume monitors, but it cannot delete them. Deleting stays in the dashboard.

Is my data safe?

The assistant uses an API key that you create and can delete at any time. It can only reach your organization, or one project if the key is limited to it, and header values and request bodies are masked in its output.

Related

Start monitoring in a few minutes

Start monitoring free

Free plan · 50 monitors · No credit card required