MCP server

JAMP exposes its monitoring through the Model Context Protocol, supported by Claude, Cursor, Claude Code and other MCP clients. An agent can read your data, and with write access it can also set JAMP up on a project and mark errors fixed. Write access stops there: an agent cannot delete sites, change billing or touch your account.

Connect with OAuth (easiest)

For clients that support OAuth, add the hosted server URL and the client sends you to JAMP to sign in and authorize. There is nothing to install and no token to paste.

https://mcp.jamp.io/mcp

JAMP is an OAuth 2.1 server with dynamic client registration and PKCE, so the client registers itself. You will see a consent screen spelling out what the agent can do, then the connection completes. No JAMP account yet? The same flow lets you create one on the way through.

Let the agent set JAMP up for you

Once connected, you can hand the whole install to the agent. Tell it "add analytics and error tracking to this site with JAMP" and it will call jamp_create_site to register the site, put the returned snippet in your layout, deploy, then poll jamp_verify_install until data arrives. You approve the connection once in the browser; the agent does the rest.

Connect with a token

For clients without OAuth, create a token under Settings → API & MCP access and pass it in an Authorization header. With Claude Code:

claude mcp add --transport http jamp \
  https://mcp.jamp.io/mcp \
  --header "Authorization: Bearer jamp_..."

For Cursor, Claude Desktop, Windsurf and similar, add the server to the MCP config:

{
  "mcpServers": {
    "jamp": {
      "url": "https://mcp.jamp.io/mcp",
      "headers": { "Authorization": "Bearer jamp_..." }
    }
  }
}

Run it locally (stdio)

If you would rather run the server yourself, the jamp-mcp package on npm is a small stdio server that wraps the same API.

claude mcp add jamp --env JAMP_API_TOKEN=jamp_... -- npx -y jamp-mcp

Tools

Most tools take a site id (from jamp_list_sites) and an optional window of 24h, 7d (default), 30d or 90d.

Alerts that wake your agent

Every alert JAMP sends (downtime, a new error, an SSL expiry, a performance regression) can also POST JSON to a webhook you set under Alerts. The payload carries a machine-readable kind and the identifiers needed to act, so you can point it at anything that starts an agent: a CI workflow, a queue worker, or a trigger that launches your coding agent with the context already filled in.

{
  "severity": "warning",
  "kind": "new_error",
  "title": "New error on my-site",
  "message": "TypeError: Cannot read properties of undefined...",
  "site": { "id": "cm...", "name": "my-site", "domain": "example.com" },
  "url": "https://jamp.io/dashboard/errors",
  "context": { "events": [{ "kind": "new_error", "key": "error:a1b2c3" }] },
  "timestamp": "2026-07-23T09:00:00.000Z"
}

The loop this enables: JAMP detects the problem, the webhook starts your agent, the agent reads the detail over MCP (jamp_get_error, jamp_uptime), fixes the code, ships, and calls jamp_resolve_error. Kinds: uptime_down, uptime_recovered, heartbeat_missed, ssl_expiring, new_error, error_spike, vitals_regression, traffic_drop, site_audit, test.

REST API

The MCP server is a thin wrapper over a read-only REST API you can also call directly with the same token. All responses are JSON and take an optional ?window= query parameter.

curl -H "Authorization: Bearer jamp_..." \
  https://jamp.io/api/v1/sites