mcpField
H
Hermes Grocy MCP

Hermes Grocy MCP

Local
github.comProductivity4Updated 13w ago

About Hermes Grocy MCP

Hermes Grocy MCP connects MCP clients to Grocy, a self-hosted household ERP, through a documented set of household inventory and shopping tools. It supports both read operations and direct changes to stock, lists, chores, tasks, and generic entities.

Key features

  1. 01Inspect products, stock, and stock entries
  2. 02Search or look up products by barcode
  3. 03Manage shopping list items
  4. 04Add, consume, and inventory stock
  5. 05Read and modify chores, tasks, and entities

Use cases

  1. 01Check due, expired, or missing products
  2. 02Update household stock after purchases or use
  3. 03Add products to a shopping list
  4. 04Review chores and mark them complete

MCP support

Exposes Grocy status, product, stock, shopping list, chore, task, and entity operations as MCP tools. Write-capable tools mutate Grocy data and private or write access requires GROCY_API_KEY.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"rusty4444-hermes-grocy": {
"command": "uvx",
"args": [
"hermes-grocy-mcp"
],
"env": {
"GROCY_BASE_URL": "<GROCY_BASE_URL>",
"GROCY_API_KEY": "<GROCY_API_KEY>"
}
}
}
}
Run commanduvx hermes-grocy-mcp
Environment variables to set
GROCY_BASE_URLGrocy base URL, with or without /api
GROCY_API_KEYSecretGrocy API key for private instances or write access

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
pypihermes-grocy-mcp0.2.1

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

Grocy MCP

A Model Context Protocol (MCP) server for Grocy, the self-hosted household ERP for groceries, inventory, chores, batteries, recipes, tasks, and shopping lists.

This server focuses on AI-friendly household operations that are awkward through generic REST clients:

  • Inspect Grocy system/version status
  • List, search, and inspect products
  • Read current stock, volatile stock, product stock details, and individual stock entries
  • List current shopping list items and add/remove products from shopping lists
  • Look up products by barcode/Grocycode
  • List and inspect any /api/objects/{entity} entity
  • Create/update generic entity objects from JSON
  • Add, consume, and inventory product stock
  • Add/remove product amounts from shopping lists

Why this exists

Grocy has a strong REST API, but MCP coverage is sparse and usually either incomplete or tightly coupled to one client's workflow. This package gives Hermes, Claude Desktop, Cursor, and other MCP clients a small, explicit, documented tool surface.

Installation

pipx install git+https://github.com/rusty4444/grocy-mcp.git

Or from a checkout:

python -m venv .venv
source .venv/bin/activate
pip install -e .

Configuration

The server reads configuration from environment variables:

VariableRequiredDefaultDescription
GROCY_BASE_URLNohttps://demo.grocy.infoGrocy base URL, with or without /api
GROCY_API_KEYNo for public/demo read-only instances, yes for private/write accessunsetGrocy API key sent as GROCY-API-KEY
GROCY_TIMEOUTNo20HTTP timeout in seconds

Grocy API keys are managed in Grocy under Manage API keys. The API accepts the GROCY-API-KEY header.

MCP client config

{
  "mcpServers": {
    "grocy": {
      "command": "grocy-mcp",
      "env": {
        "GROCY_BASE_URL": "https://grocy.example.com",
        "GROCY_API_KEY": "your-api-key"
      }
    }
  }
}

Tools

ToolPurpose
grocy_system_infoGrocy version and runtime details
grocy_list_productsList configured products
grocy_search_productsSearch products by name/description
grocy_get_productFetch one product object
grocy_lookup_product_by_barcodeResolve a barcode/Grocycode
grocy_stock_overviewCurrent stock rows
grocy_volatile_stockDue, overdue, expired, or missing products
grocy_product_stock_detailsDetailed stock state for one product
grocy_product_stock_entriesIndividual stock entries in next-use order
grocy_common_entitiesCommon generic entity names useful with CRUD tools
grocy_list_shopping_listsConfigured shopping lists
grocy_list_shopping_list_itemsCurrent shopping list rows, optionally filtered by list id
grocy_list_entityList any generic Grocy entity
grocy_get_entity_objectFetch any generic entity object
grocy_create_entity_objectPOST a generic entity object from JSON
grocy_update_entity_objectPUT a generic entity object from JSON
grocy_add_stockAdd product amount to stock
grocy_consume_stockConsume/remove product amount from stock
grocy_inventory_productSet product inventory amount
grocy_add_product_to_shopping_listAdd a product to a shopping list
grocy_remove_product_from_shopping_listRemove a product from a shopping list
grocy_set_userfieldsSet userfield values on an entity object
grocy_list_choresList chores including computed next-due times
grocy_execute_choreExecute (mark done) a chore, optionally backdated
grocy_list_tasksList tasks

Development and validation

python -m venv .venv
source .venv/bin/activate
pip install -e '.[dev]'
ruff check .
pytest
python scripts/live_readonly_test.py

The live read-only test defaults to https://demo.grocy.info, avoiding mutations on shared infrastructure. It has been validated against Grocy API 4.6.0. Use a private Grocy instance plus GROCY_API_KEY for write-path testing.

Safety

Write-capable tools directly mutate Grocy data. Prefer read-only tools when using public demos. Keep GROCY_API_KEY in MCP client environment config or a secret manager, never in source control.

Comments

Markdown supported

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.