mason-levyy/myfitnesspal-mcp — explained in plain English
Analysis updated 2026-05-18
Log meals to your MyFitnessPal diary by telling an AI assistant what you ate.
Search the food database for macros before logging the exact item.
Track weight, calories, or protein trends over a date range from a chat conversation.
Export your full nutrition history for analysis.
| mason-levyy/myfitnesspal-mcp | 0marildo/imago | 0xdfi/glm-5.2-1m-4x-dgx-spark | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | easy | hard |
| Complexity | 3/5 | 2/5 | 5/5 |
| Audience | general | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires copying a session cookie from your browser after logging into MyFitnessPal manually.
This project connects the food tracking app MyFitnessPal to Claude or any other AI assistant that supports the Model Context Protocol, so you can log meals, look up foods, and check your nutrition trends just by talking to the assistant instead of opening the app. The README explains this is unofficial, since MyFitnessPal has no public way for outside apps to connect. Instead, it works by mimicking what the MyFitnessPal website itself does in a browser. MyFitnessPal recently added extra security that broke the simple username and password logins other tools used to rely on, so this project instead asks you to copy a login cookie out of your browser after logging in normally. That cookie is used to authenticate, disguised so it looks like it is coming from a real Chrome browser rather than a script. There is also an optional feature that automatically refreshes that login using a hidden browser window when the session expires, so you do not have to repeat the copy and paste process every month. Once set up, the assistant gains a set of tools: it can read a full day's food log, search the food database for matches with calorie and nutrient details, add a specific food to your diary, delete or change an entry, log your weight, read and write the notes section of your diary, save a personal note about how you feel that stays only on your own computer, and pull trends like weight or protein intake over a date range, or export everything at once. To log something accurately, you first search for the food, then tell the assistant to log the exact match rather than trusting a guess. Reading water intake works, but MyFitnessPal does not currently offer a way to write it back through this method. The tool keeps a local database that fills in missing days automatically the first time you use it. It can also run as a small local web server instead of the default mode, though the README is clear that mode has no built-in login protection and should never be exposed directly to the internet without something else guarding it. Setup requires a tool called uv to install and run it. The project is released under the MIT license.
Lets Claude or other AI assistants log meals, search foods, and track nutrition trends in your real MyFitnessPal account.
Mainly Python. The stack also includes Python, MCP, SQLite.
Use, modify, and share freely for any purpose, including commercial use, as long as you keep the copyright notice.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.