Uptime Agent

You connect your Uptime Agent monitoring system to AI assistants using MCP to query uptime status, view incidents, and create monitors through natural conversations. This enables real-time insights and streamlined management of your monitoring data via AI prompts.

How to use

Interact with your MCP server through your chosen MCP client to check monitors, review incidents, and create new monitoring endpoints. You can ask the assistant to show all active monitors, retrieve details for a specific monitor, or set up a new monitor for a given URL. When you describe incidents, you can ask questions like which incidents occurred recently and what the downtime looked like, all through natural language.

Common workflows you can perform include: checking all active uptime monitors, creating a new monitor for a URL or API endpoint, listing incidents by monitor or over a time range, and retrieving incident details for deeper analysis.

How to install

Prerequisites you need before installing: Node.js 18 or higher, an active Uptime Agent account, and your Uptime Agent API key.

Step 1. Quick install via npm (recommended) wants you to run a single command that installs the MCP server and configures it for Claude Desktop, then prompts you for your API key and completes setup.

Command: npx uptime-agent-mcp setup

Step 2. Install via Smithery.ai (alternative) requires you to install using a provided CLI and client key from your Smithery account.

Command: npx -y @smithery/cli@latest install @AVIMBU/uptime_agent_mcp --client claude --key <personal_key>

Step 3. Manual local installation (advanced users) lets you clone the project, install dependencies, and build before running.

Commands to run locally:

Configure with your API key by creating a .env file at the project root with:

UPTIME_API_KEY=your-api-key-here
PORT=3000  # Optional, defaults to 3000

Run the MCP server

Start the server once configuration is in place. You can start from the built distribution or from the source, depending on your setup.

Start command (example):

npm start
# or if you built the project
node dist/index.js

Configuration and security

Security is important when exposing monitoring data through AI assistants. Use the MCP integration with a strong API key, limit access to trusted clients, and monitor usage patterns to detect unusual activity.

Environment variable shown for integration:

UPTIME_API_KEY=YOUR_API_KEY

Available tools

listMonitors

Get a complete list of all monitoring endpoints

getMonitor

Retrieve detailed information about a specific monitor using its id

createMonitor

Set up a new endpoint to monitor with a name, URL, tracking type, and check frequency

listIncidents

View all detected downtime incidents

getIncident

Get detailed information about a specific incident by its id

listIncidentsByMonitor

See all incidents for a particular endpoint using its monitor_id