nutrition-mcp
Local-first nutrition tracker MCP server for Hermes, enabling food, alias, recipe, and meal log management with SQLite persistence.
README
nutrition-mcp
A small local-first nutrition tracker exposed as an HTTP MCP server for Hermes.
The database is the source of truth. Hermes should call tools here to remember foods, aliases, recipes, and meal logs instead of storing nutrition facts in agent memory.
Design
The server stores three useful concepts:
- Foods: atomic macro sources such as Babybel, Greek yogurt, tuna, dough, mozzarella, or homemade sauce.
- Aliases: stable names Hermes can resolve, such as
babybel,default tuna, orpizza dough. - Recipes: reusable templates composed from foods, such as
tuna pizza = 165g dough + 60g tuna + 40g cheese.
Recipes can be logged with one-off adjustments. For example, after creating a tuna pizza recipe, Hermes can log it with:
{
"alias": "tuna pizza",
"adjustments": [
{"alias": "cheese", "delta_grams": 20}
]
}
The logged meal stores macro snapshots and a component snapshot, so historical logs do not change when you later edit a food or recipe.
Tracked nutrients are calories, protein, carbs, fat, fiber, sugars, saturated fat, and salt. Salt is stored as grams of salt from nutrition labels, not sodium in milligrams.
MCP Endpoint
- URL:
http://HOST:8765/mcp - Transport: Streamable HTTP via the official Python MCP SDK v1
FastMCP - Health endpoints:
GET /GET /health
If MCP_TOKEN is set, every HTTP request must include:
Authorization: Bearer <token>
Tools
Core tools:
add_foodupdate_foodadd_aliasget_foodsearch_foodslist_foodslog_foodget_dayget_entriesdelete_entrydelete_foodupdate_entryfinalize_dayget_weekly_reportlist_aliaseshealth
Recipe tools:
add_recipeupdate_recipeget_recipesearch_recipeslog_recipedelete_recipe
Data Paths
Inside the container:
- SQLite DB:
/data/nutrition.db - Exports:
/data/exports - Daily Markdown:
/data/exports/daily/YYYY-MM-DD.md - CSV:
/data/exports/csv/YYYY-MM-DD.csv - JSON:
/data/exports/json/YYYY-MM-DD.json
Local Development
py -m venv .venv
.\.venv\Scripts\python -m pip install -e ".[test]"
.\.venv\Scripts\pytest
.\.venv\Scripts\python -m app.main
On Linux/macOS:
python -m venv .venv
. .venv/bin/activate
python -m pip install -e ".[test]"
pytest
python -m app.main
Test the health endpoint:
curl http://localhost:8765/health
With a token:
curl -H "Authorization: Bearer change-me" http://localhost:8765/health
Demo Data
After the server is running, add demo foods through MCP:
{
"name": "Babybel",
"serving_name": "1 piece",
"kcal": 70,
"protein_g": 4,
"carbs_g": 0,
"fat_g": 5,
"fiber_g": 0,
"sugars_g": 0,
"saturated_fat_g": 0,
"salt_g": 0,
"aliases": ["babybel"]
}
{
"name": "Greek yogurt",
"serving_name": "1 serving",
"kcal": 180,
"protein_g": 20,
"carbs_g": 8,
"fat_g": 4,
"fiber_g": 0,
"sugars_g": 0,
"saturated_fat_g": 0,
"salt_g": 0,
"aliases": ["greek yogurt"]
}
Example recipe ingredients use grams when the food has grams_per_serving:
{
"name": "Tuna pizza",
"aliases": ["tuna pizza"],
"ingredients": [
{"alias": "pizza dough", "grams": 165},
{"alias": "default tuna", "grams": 60},
{"alias": "cheese", "grams": 40}
]
}
Docker
Build:
docker build -t nutrition-mcp:0.3.0 .
Run:
docker run -d \
--name nutrition-mcp \
-p 8765:8765 \
-e TZ=Europe/Bucharest \
-e MCP_TOKEN=change-me \
-e PUBLIC_HOSTS=192.168.1.142,nutrition-mcp \
-v /mnt/user/appdata/nutrition-mcp:/data \
nutrition-mcp:0.3.0
For a published image:
docker run -d \
--name nutrition-mcp \
-p 8765:8765 \
-e TZ=Europe/Bucharest \
-e MCP_TOKEN=change-me \
-e PUBLIC_HOSTS=192.168.1.142,nutrition-mcp \
-v /mnt/user/appdata/nutrition-mcp:/data \
ghcr.io/REPLACE_ME/nutrition-mcp:0.3.0
Unraid Add Container
- Name:
nutrition-mcp - Repository:
ghcr.io/REPLACE_ME/nutrition-mcp:0.3.0 - Network Type:
bridge - Port: host
8765-> container8765TCP - Path:
/mnt/user/appdata/nutrition-mcp->/data - Env:
TZ=Europe/BucharestMCP_TOKEN=<long random token>PUBLIC_HOSTS=192.168.1.142,nutrition-mcp
- WebUI:
http://[IP]:[PORT:8765]/
Hermes MCP Config
If Hermes connects over LAN:
mcp_servers:
nutrition:
url: "http://192.168.1.142:8765/mcp"
headers:
Authorization: "Bearer change-me"
If Hermes and this server are on the same custom Docker network:
mcp_servers:
nutrition:
url: "http://nutrition-mcp:8765/mcp"
headers:
Authorization: "Bearer change-me"
Environment
DATA_DIR=/data
TZ=Europe/Bucharest
MCP_TOKEN=
PUBLIC_HOSTS=192.168.1.142,nutrition-mcp
HOST=0.0.0.0
PORT=8765
Notes
- SQLite is the only database.
- All persistent files live under
/data. - No arbitrary SQL tool is exposed.
- Deletes require an exact meal
entry_id. - Historical entries keep snapshots when foods or recipes are updated.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.