DataFrame MCP server

Connect Claude, Cursor, GPT-4, Gemini, Grok, DeepSeek, or another assistant to your training. The assistant opens DataFrame so you can sign in and choose what it may do, then you return to the chat. Do not ask the athlete to paste an access token into chat.

Add DataFrame as an MCP server for Claude, Cursor, GPT-4, Gemini, Grok, DeepSeek, or another AI assistant. Sign in on DataFrame, choose what it may view or change, then return to the chat.

Configuration

{
  "mcpServers": {
    "DataFrame.fit": {
      "url": "https://dataframe.fit/api/mcp"
    }
  }
}

Server URL

https://dataframe.fit/api/mcp

Plain-text notes for assistants: https://dataframe.fit/mcp.md

How to connect

Use the configuration above, or add a remote MCP server with that URL in Claude, Cursor, GPT-4, Gemini, Grok, DeepSeek, or another assistant.

What you can allow

Activities, training plan, nutrition, wellness, races, calendar, gear, and preferences. Viewing starts on. Changes, and push to Garmin, stay off until you tick them. Agents never get admin access.

You can connect more than one assistant. Each one has its own access. Revoke one under Settings without affecting the others.

Authorize screen with View and Create or update columns. Viewing is selected. Changes stay off.
You choose what that assistant may view or change. Viewing starts on. Create or update, and push to Garmin, stay off until you tick them.
Cursor tools list showing DataFrame.fit connected, with 23 tools and 2 resources enabled.
After you allow it, the assistant lists DataFrame.fit as connected.

mcp.md (plain text for assistants) · Step-by-step guide · What each permission does

More about DataFrame.fit

© 2026 DataFrame™. All rights reserved.

This project is currently fully supported by Scope (www.scope.mt).

DataFrame™ is a non-commercial personal project. It is not affiliated with, endorsed by, or connected to Garmin, TrainingPeaks, Peaksware, or other fitness brands.