Guides / Claude Desktop MCP
Geomanic provides an MCP server so AI assistants can work with your travel data. Once connected, you can ask Claude Desktop things like โHow many days did I spend in Spain last year?โ or โWhat was my longest trip in June?โ โ and it answers from your real GPS data. Setup takes a few minutes.
Log in to Geomanic and open the Data page. In the MCP Integration tile, generate your personal MCP API key. Treat it like a password โ you can regenerate it there at any time if it is ever compromised.
The easiest way is the ready-made skill:
Alternatively, download the skill template from the geomanic-mcp repository (skills/claude/SKILL.md) and replace YOUR_API_KEY with your actual key.
/skills install @weltspion/geomanic, then set your key with /secrets set GEOMANIC_TOKEN YOUR_MCP_KEY.bridge/config/cursor/mcp.json).https://geomanic.com/api/v1/mcp with the header Authorization: Bearer YOUR_MCP_KEY.The Geomanic MCP server exposes these tools:
Full parameter documentation is available in the repository docs.
The Model Context Protocol (MCP) is an open standard that lets AI assistants like Claude Desktop connect to external services. With the Geomanic MCP server, Claude can read and manage your GPS data on your behalf โ you simply ask in natural language.
Treat the key like a password: never share it or commit it to git. The key only grants access to your own Geomanic data, and you can revoke and regenerate it at any time on the Data page.
Claude Desktop (via skill), OpenClaw (via ClawHub skill), Cursor (via the Node.js bridge), and any MCP client that supports remote servers with a URL and bearer token.
No GPS data yet? Start with the Overland or OwnTracks setup guide.
Ready to start? Create a free Geomanic account.