About Memory.fit
Memory.fit is a hosted MCP server that connects an AI to a person’s training, sleep, recovery, and nutrition data. It lets the AI read connected activity and health records, then write permitted plans and logs to the Memory.fit app.
Key features
- 01Read workouts, sleep, HRV, and nutrition
- 02Access race goals and body composition
- 03Write training sessions and daily targets
- 04Create meal plans, meal logs, and notes
- 05Retrieve updates since the last check
Use cases
- 01Review a week of training and recovery data
- 02Draft upcoming sessions for a connected watch
- 03Set daily nutrition targets and meal plans
MCP support
It provides a hosted Streamable HTTP MCP server at `https://api.memory.fit/mcp` for AI clients. Authentication uses OAuth 2.1 or a personal bearer key, and users authorize permitted access and writes.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://api.memory.fit/mcpclaude mcp add --transport http fit-memory-memory-fit https://api.memory.fit/mcp{"mcpServers":{"fit-memory-memory-fit":{"type":"http","url":"https://api.memory.fit/mcp"}}}Open in Cursor{"servers":{"fit-memory-memory-fit":{"type":"http","url":"https://api.memory.fit/mcp"}}}Open in VS Code{"mcpServers":{"fit-memory-memory-fit":{"type":"http","url":"https://api.memory.fit/mcp"}}}{"mcpServers":{"fit-memory-memory-fit":{"serverUrl":"https://api.memory.fit/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEMemory.fit MCP
Memory.fit keeps your training, sleep, recovery and fuel in one place for the AI you already use. Your AI reads workouts from your watch and apps, sleep and HRV, and what you ate, then writes your week back: sessions, daily targets and meal plans, which show in the Memory.fit app on iPhone and Android.
This repository holds only the connection settings for Memory.fit MCP, a hosted server. There is nothing to run or build.
- Address:
https://api.memory.fit/mcp(Streamable HTTP) - Sign-in: OAuth 2.1 with dynamic client registration. Your AI opens the Memory.fit sign-in page; you choose what it may do. No key needed. AIs without sign-in can use a personal key from the app (Settings › AI agents › Use a key instead), sent as
Authorization: Bearer mem_pat_…. - Needs: a Memory.fit account, set up in the app first.
Connect
| App | How |
|---|---|
| Claude | Settings › Connectors › Add custom connector, paste the address |
| ChatGPT | Settings › Apps and connectors › Create, paste the address |
| Grok, Grok Bot | Connectors › New connector › Custom, paste the address |
| Cursor | Install this plugin, or add mcp.json from this repo to ~/.cursor/mcp.json |
| Claude Code | claude mcp add --transport http memory-fit https://api.memory.fit/mcp |
| Codex | codex mcp add memory-fit --url https://api.memory.fit/mcp |
| Gemini CLI | gemini extensions install https://github.com/Fun-Management-SRL/memory-fit-mcp |
| Cline | MCP Servers › Remote Servers: name memory-fit, URL https://api.memory.fit/mcp, transport Streamable HTTP (or the JSON below) |
| Anything else | Any MCP client: the address above, then sign in |
Step-by-step pages for each AI: memory.fit/ai-coaches.
Cline, or any client with a JSON settings file
{
"mcpServers": {
"memory-fit": {
"type": "streamableHttp",
"url": "https://api.memory.fit/mcp",
"disabled": false
}
}
}
The first call opens the Memory.fit sign-in (OAuth 2.1). A client without OAuth can use a personal key from the app (Settings › AI agents › Use a key instead) as a header:
"headers": { "Authorization": "Bearer mem_pat_..." }
What your AI can do
- Read: workouts (Garmin, Wahoo, COROS, Polar, Suunto, Apple Health, Health Connect and more), sleep, HRV and resting heart rate, body composition, what you ate and drank, your races and goals.
- Write, with your permission: training sessions (sent to your watch), daily targets, meal plans, logged meals, notes for your other AIs.
- Check in:
get_updatesreturns only what arrived since it last looked.
Memory.fit recommends nothing itself: the plans are your AI's, the data is yours. It is not medical advice.
Links
- Website: https://memory.fit
- Privacy: https://memory.fit/privacy
- Terms: https://memory.fit/terms
- Support: [email protected]
The connection settings in this repository are MIT licensed. Memory.fit, the server and the app are not.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.