Let your AI agent run your monitoring
Uptrack ships a hosted MCP server. Connect Claude or ChatGPT, and it can create monitors, check on incidents, and run real-browser checks for you — in plain language, on your account. No scripts, no separate dashboard to babysit.
https://api.uptrack.app/mcpIt authenticates with OAuth on first use — approve once and your agent is connected. Tool access is scoped to your organization, same as the dashboard.
Connect over MCP
Add Uptrack as a connector in Claude or ChatGPT once. It speaks the Model Context Protocol over Streamable HTTP, with OAuth — no API keys to paste, no glue code to write.
Ask, don’t click
"Watch app.example.com and tell me if checkout breaks." Your agent creates the monitor, lists open incidents, acknowledges one to pause escalation, or pulls uptime — through real tools, scoped to your account.
Run a browser check from a chat
There’s a public, no-signup browser-check tool an agent can call directly: give it a URL, get back a screenshot, Core Web Vitals, and a verdict. Pages on this site also expose a WebMCP tool the browser can use.
Built to be read by machines
Uptrack scores Level 5 "Agent-Native" on isitagentready.com — an MCP server card, an A2A agent card, a discovery doc at /auth.md, an api-catalog, and AI rules in robots.txt. Agents can find and use it without a human in the loop.
Start free — 10 monitors at 30-second checks
Create an account, then connect your agent. Or see Uptrack’s agent-readiness score.