UpCheck

MCP server for uptime monitoring

Let any AI agent create, check, and manage your monitors itself.

Your monitors, in your agent's hands

UpCheck ships a remote MCP server, so any Model Context Protocol client, Claude or your own agent, can manage monitors on your behalf. Say "monitor my site" and the agent drives the rest.

The server exposes five tools: list_monitors for current status and 24h stats, create_monitor (only a URL required), update_monitor (including enabled: false to pause), delete_monitor, and test_monitor to run one check against a draft config without saving. It speaks stateless streamable HTTP, which every current MCP client supports; there is nothing to install or host.

claude mcp add --transport http upcheck \
https://upcheck-api-a.rekwiem.com/mcp \
--header "Authorization: Bearer upk_your_key"

Agents that set up their own monitoring

The interesting flow isn't you asking an agent to make a monitor. It's the agent doing it unprompted: it deploys a service, calls test_monitor to confirm the endpoint answers, then create_monitor to register it, and the thing it just shipped is being watched before the deploy log scrolls away.

Because create returns the first check's result, the agent knows immediately whether it pointed the monitor at the right URL, and can correct itself without a human in the loop.

Scoped to one project, by construction

Authentication is a Bearer project API key, created in your project's Settings under Access tokens. A key binds to one project, so the agent can only touch that project's monitors; it can never create more keys or reach account settings.

Give the key an expiry for throwaway experiments, and revoke it any time from the same list. In clients with a connector UI, paste the endpoint and set the Authorization header; for clients configured by file, the JSON above is the whole setup.

More for developers

Read more

MCP Server questions

Any client that speaks the Model Context Protocol over streamable HTTP, which every current client does: Claude, Claude Code, or an agent you build yourself. Add it as a remote server with one URL and an Authorization header.

Five: list_monitors, create_monitor, update_monitor, delete_monitor, and test_monitor. Between them an agent can see status and 24h stats, register new monitors, pause or edit existing ones, and dry-run a config without saving it.

No. MCP access is included on every plan, with the same project API keys as the REST API. Plan limits still apply: a create that fails with monitor_limit means you've hit your monitor cap.

One project's monitors, nothing more. The key it authenticates with is project-scoped, can't mint more keys, and can't reach account settings. You can give it an expiry and revoke it from Settings at any time.

Set it up in minutes, never miss an outage again.

Free to start · One-minute checks · iPhone, iPad, and Mac