MCP Integration

Ask your AI assistant about your monitors

Connect exit1.dev to Claude, Cursor, or Windsurf via the Model Context Protocol. Check uptime, investigate failures, and compare response times — all in natural language. No dashboards required.

Set Up MCPSign In

Key Features

Everything you need to monitor your infrastructure effectively

Natural Language Queries

Ask "are any of my monitors down?" or "what's the uptime for my API this week?" and get instant answers. No dashboards, no clicking, no context switching.

Works with Claude, Cursor & Windsurf

Connect to Claude Desktop, Claude Code, Cursor, or Windsurf. Any AI tool that supports the Model Context Protocol can plug in.

5 Read-Only Tools

List checks, get check details, pull historical results, query aggregate stats, and view status pages. All read-only — your data stays safe.

One-Line Setup

Install via npx with zero dependencies. Add your API key to the config and restart your AI assistant. Connected in under 2 minutes.

Uptime & Performance Stats

Query uptime percentages, average/min/max response times across multiple time ranges. Compare this week vs last week in a single prompt.

Incident Investigation

Ask your AI assistant to show recent failures, filter by status, and dig into historical check results with timestamps and error details.

Why Choose exit1.dev?

See how we stack up against the competition

Featureexit1.devOthers
Natural language monitoring queries
Claude Desktop support
Cursor IDE support
Windsurf support
Claude Code (CLI) support
Read-only (safe) accessN/A
Zero-dependency install (npx)N/A
Historical data queries
Multi-range stat comparison
Available at $3/moEnterprise only

Technical Details

Built for developers, by developers

Architecture

The exit1-mcp server runs locally on your machine as a stdio-based MCP server. It communicates with your AI assistant through the Model Context Protocol standard. All API calls go directly from your machine to the Exit1 API — no intermediate servers, no data stored locally.

Performance

Tool calls execute in under 500ms for most queries. The server starts instantly via npx with zero warm-up time. Rate limits are enforced server-side at 5 requests/minute per API key, with generous daily limits for normal AI conversation patterns.

API

The MCP server exposes 5 tools: list_checks, get_check, get_check_history, get_check_stats, and get_status_page. All tools are read-only and require a checks:read scoped API key. The server is published as exit1-mcp on npm and distributed via npx.

Frequently Asked Questions

Everything you need to know about our monitoring

MCP (Model Context Protocol) is an open standard that lets AI assistants connect to external tools and data sources. Exit1's MCP server gives your AI assistant read-only access to your monitoring data, so you can ask questions in natural language instead of navigating dashboards.

Claude Desktop, Claude Code, Cursor, and Windsurf are officially supported with setup instructions. Any AI tool that implements the Model Context Protocol can connect using the same npx-based configuration.

Yes. The MCP server provides read-only access. It can list checks, get details, pull history, and view stats — but it cannot create, modify, or delete anything. Your API key needs only the checks:read scope.

Create an API key in your Exit1 dashboard with checks:read scope. Then add the exit1 MCP server config to your AI assistant (a JSON block with your API key). The whole process takes under 2 minutes. See the full setup guide in our docs.

Anything about your monitoring data. Examples: "Are any monitors down?", "What's the uptime for my API over 30 days?", "Show the last 10 failures for production", "Compare response times this week vs last week", "What does my status page show right now?"

The MCP server uses your existing API rate limits: 5 requests/minute per key, 500 requests/day per key, and 2,000 requests/day per user. A typical AI conversation uses 2-5 tool calls, so you're unlikely to hit these limits.

MCP is a Nano plan feature. Free plan users can upgrade to Nano starting at $3/month to access it, along with unlimited monitors, SMS alerts, and all other Nano features.

You need Node.js 18+ installed on your machine. The MCP server itself installs automatically via npx — no manual package installation required.

MCP Is a Nano Feature

Connect your AI assistant to exit1.dev and query monitoring data conversationally. Available on the Nano plan starting at $3/month.

Get Nano — $3/moLearn More

API & Webhooks

Full REST API access for programmatic integration. The MCP server is built on top of the same API.

Analytics & Reports

Track uptime trends and response times. MCP lets you query the same data conversationally.

Smart Alerting

Get notified via email, SMS, and webhooks. Use MCP to investigate after alerts fire.

Status Pages

Public status pages for your customers. Query their current state via MCP.

Real-Time Monitoring

Live status updates for websites and APIs. Ask your AI assistant about current status anytime.

Global Monitoring

Monitor from multiple regions. MCP surfaces data from all regions in a single conversation.

Start Monitoring