mealie-mcp

mealie-mcp

Exposes Mealie recipe manager as LLM-callable tools for searching recipes, managing meal plans, and editing shopping lists.

Category
访问服务器

README

mealie-mcp

A Model Context Protocol server that exposes Mealie — a self-hosted recipe manager — as a set of LLM-callable tools. It lets Claude Desktop / Claude Code (or any MCP client) search recipes, fetch details, manage the meal plan, and edit shopping lists on your own Mealie instance.

Tools

Tool Purpose
search_recipes(query?, tags?, limit?) Search recipes; returns slug, name, description, tags, categories
get_recipe(slug) Full recipe JSON for a slug
list_meal_plan(start_date, end_date) Meal plan entries between two ISO dates
list_shopping_lists() IDs and names of all shopping lists
add_shopping_list_items(list_id, items[]) Add free-text items to a list
create_recipe(name) Create a blank recipe; returns generated slug
create_meal_plan_entry(date, entry_type, recipe_slug?, title?) Add a meal plan entry

Requirements

  • Python 3.11+
  • A running Mealie instance (self-hosted; tested against Mealie v2)
  • A long-lived Mealie API token (User Profile → API Tokens in the Mealie UI)

Configuration

Copy .env.example to .env and fill in the values:

cp .env.example .env
Variable Required Default Description
MEALIE_URL yes — Base URL of your Mealie instance (e.g. http://localhost:9011 or http://mealie:9000)
MEALIE_API_TOKEN yes — Long-lived bearer token from Mealie
MCP_TRANSPORT no sse stdio for local subprocess, sse for HTTP streaming
MCP_HOST no 0.0.0.0 Bind address for SSE mode
MCP_PORT no 8000 Bind port for SSE mode

Install & run locally

pip install -e .
# stdio mode (for Claude Desktop):
MCP_TRANSPORT=stdio mealie-mcp
# sse/http mode (for docker-compose / remote clients):
MCP_TRANSPORT=sse MCP_PORT=8000 mealie-mcp

The SSE endpoint is served at http://<host>:<port>/sse.

Claude Desktop configuration

Edit Claude Desktop's config file (macOS: ~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "mealie": {
      "command": "mealie-mcp",
      "env": {
        "MEALIE_URL": "http://localhost:9011",
        "MEALIE_API_TOKEN": "paste-your-token-here",
        "MCP_TRANSPORT": "stdio"
      }
    }
  }
}

If mealie-mcp is not on your PATH, point command at the full binary path (e.g. /Users/you/.venvs/mealie-mcp/bin/mealie-mcp) or invoke via Python:

{
  "mcpServers": {
    "mealie": {
      "command": "python",
      "args": ["-m", "mealie_mcp"],
      "env": {
        "MEALIE_URL": "http://localhost:9011",
        "MEALIE_API_TOKEN": "paste-your-token-here",
        "MCP_TRANSPORT": "stdio"
      }
    }
  }
}

Restart Claude Desktop after editing the config.

Docker / docker-compose

A Dockerfile is included. To run standalone:

docker build -t mealie-mcp .
docker run --rm -p 8765:8000 \
  -e MEALIE_URL=http://host.docker.internal:9011 \
  -e MEALIE_API_TOKEN=your-token \
  mealie-mcp

To add the server to an existing Mealie stack, see docker-compose.snippet.yml:

services:
  mealie-mcp:
    build:
      context: ./mealie-mcp
    depends_on:
      - mealie
    environment:
      MEALIE_URL: "http://mealie:9000"
      MEALIE_API_TOKEN: "${MEALIE_API_TOKEN}"
      MCP_TRANSPORT: "sse"
      MCP_HOST: "0.0.0.0"
      MCP_PORT: "8000"
    ports:
      - "8765:8000"

Point your MCP client at http://<docker-host>:8765/sse.

Testing the server manually

With the server running in SSE mode, confirm it is reachable:

curl -N http://localhost:8000/sse

You should see an SSE event stream open. For stdio mode, Claude Desktop handles the handshake — there is no HTTP endpoint.

Project layout

mealie-mcp/
├── pyproject.toml
├── Dockerfile
├── docker-compose.snippet.yml
├── .env.example
├── README.md
└── src/mealie_mcp/
    ├── __init__.py
    ├── __main__.py        # CLI entry point (loads .env, dispatches transport)
    ├── server.py          # FastMCP server + tool definitions
    └── client.py          # Async httpx wrapper for the Mealie REST API

License

MIT

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选