Fitrofy connector for AI assistants
Use the Fitrofy API inside Claude, Cursor, Gemini CLI and any other assistant that supports MCP (Model Context Protocol). The connector covers Indian food search, barcode, photo and label scans, and the Fitrofy diet brain: user profiles, food logging, insights, diet plans and "can I eat this?" answers.
It runs on your own computer, and calls go to Fitrofy from your computer's IP address.
Before you start
- Create a free account at https://apiportal.fitrofy.com. ₹500 is added to your account.
- Under API keys, create a key. Add this computer's public IP address to the key: every Fitrofy key accepts calls only from the addresses it lists.
- Install uv. The commands below use
uvxto run the connector.
Add it to your assistant
Claude Desktop
Settings → Developer → Edit config, then add:
{
"mcpServers": {
"fitrofy": {
"command": "uvx",
"args": ["fitrofy-mcp"],
"env": {
"FITROFY_API_KEY": "fsk_..."
}
}
}
}
Claude Code
claude mcp add fitrofy -e FITROFY_API_KEY=fsk_... -- uvx fitrofy-mcp
Cursor
Add the same mcpServers block as Claude Desktop to ~/.cursor/mcp.json.
Gemini CLI
gemini extensions install https://github.com/Fitrofy/fitrofy-gemini-extension
Then set FITROFY_API_KEY to your key's secret in your shell.
Tools
| Tool | What it does | Price (+ GST) |
|---|---|---|
food_search |
Calories, macros and health ratings for an Indian food | ₹1; ₹0.50 after 2,000 a month; no match free |
scan_barcode |
A packaged food by its barcode | ₹1; not found free |
scan_food_image |
The foods in a photo | ₹2 |
scan_label |
A nutrition label from one or two photos | ₹2 |
register_user |
A user's profile and daily targets | free |
get_user |
A user's profile and targets | free |
log_meals |
Log what a user ate, in plain words | ₹10 per user per date; the rest of that date free |
get_day_log, get_insights, get_recommendations |
The logged day, its insights, next-meal advice | free once that date's log is paid |
get_diet_plan |
A user's diet plan for a date | ₹1.50 per user per date |
ask_food_question |
"Can I eat…?" or "What should I eat…?" for a user | ₹1 |
Every paid call is charged to your Fitrofy balance. The key's hourly, daily and monthly spend caps apply, as they do to any call.
scan_food_image and scan_label also accept a file path on your computer (image_path,
image_paths).
No personal data
A user is named only by user_ref, your own pseudonymous ID (for example acme-u-7f3a9c).
Never put a person's name, email address or phone number in a request; Fitrofy refuses them.
What you type into an AI assistant is also processed under that assistant's own terms.
Fitrofy's answers are food information, not medical advice.
Settings
| Variable | |
|---|---|
FITROFY_API_KEY |
Your key's secret (fsk_…), shown once when the key is made. It is sent as the Bearer token on every call. |
FITROFY_BASE_URL |
Optional; defaults to https://apiportal.fitrofy.com |
If Fitrofy refuses the key, the connector stops calling until it is restarted, because repeated refusals lock this computer's address for 15 minutes. A key made before keys could be used directly needs rotating once on the API keys page; use the new secret.
Development
The tool list is built from the public OpenAPI (web/public/openapi/fitrofy-components.v1.yaml)
and the HTTP work is the Python client (clients/python/fitrofy.py); both are copied into the
wheel at build time. Test: python -m pytest clients/mcp/tests.
mcp-name: io.github.Fitrofy/fitrofy-mcp
Metadata
Release files for fitrofy-mcp 0.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| fitrofy_mcp-0.2.0-py3-none-any.whl | Python 3 | none | any | Details for fitrofy_mcp-0.2.0-py3-none-any.whl |
Release files / fitrofy_mcp-0.2.0-py3-none-any.whl
| Download URL | fitrofy_mcp-0.2.0-py3-none-any.whl |
|---|---|
| Size | 24.8 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
670db12b7eb2778930327387787248c8202160582f4e91dc68257407eefac739
|
|
BLAKE2b-256 checksum How to use checksums |
ba97737d17d32c1b446c84e0d3621fd4f253b2f7b7d1b14ef6f79dfc1831b480
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.5
|
Security for fitrofy-mcp 0.2.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.