MeteoSwiss MCP Server
Provides access to official MeteoSwiss weather data, including regional reports, daily forecasts, and website search functionality across multiple languages. It enables AI assistants to retrieve real-time weather information and documentation from MeteoSwiss using the Model Context Protocol.
README
(Demo) MCP Server for MeteoSwiss Data
A Model Context Protocol (MCP) server for MeteoSwiss weather data.
Overview
This server provides weather data from MeteoSwiss using the Model Context Protocol (MCP), allowing AI assistants like Claude to access weather information.
Features
- Weather reports for regions in Switzerland (North, South, West)
- Search MeteoSwiss website content across multiple topics
- Fetch full content from MeteoSwiss pages with format conversion
- Multi-language support (German, French, Italian, English)
- Weather forecasts with daily breakdowns
- Intelligent HTTP caching with ETag support
- Test fixtures for development
Architecture
This MCP server runs as an HTTP service with Server-Sent Events (SSE) for real-time communication. It's designed to be accessed remotely using mcp-remote for Claude Desktop integration.
Quick Start
This server runs as an HTTP service and can be accessed using mcp-remote:
# Start the server
npm start # or: pnpm start
# In Claude Desktop, use:
npx mcp-remote http://localhost:3000/mcp
Development Mode
For development with auto-reload:
# Start with file watching
npm run dev # or: pnpm dev
# Test with MCP Inspector
npm run dev:inspect
Development
This project uses tsx for TypeScript execution, providing a smooth development experience with hot reloading.
Prerequisites
- Node.js v18.0.0 or later (we recommend using nvm for Node.js version management)
- pnpm for package management
Setting up the development environment
-
Clone the repository:
git clone https://github.com/eins78/mcp-server-meteoswiss-data.git cd mcp-server-meteoswiss-data -
Install dependencies:
pnpm install -
Use the correct Node.js version:
nvm use
Running the application
Running the Server
The server runs as an HTTP service with Server-Sent Events (SSE):
# Start the server (default port 3000)
pnpm start
# Or specify a custom port
PORT=8080 pnpm start
The HTTP server provides:
- GET
/- Server information - GET
/mcp- MCP SSE endpoint for client connections - POST
/messages?sessionId=...- Message handling endpoint - GET
/health- Health check endpoint
Development Mode
Start the development server with hot reloading:
pnpm run dev
Or run the application without hot reloading:
pnpm run start
Type checking
Check TypeScript types without emitting JavaScript files:
pnpm run lint
Project Structure
src/- Source codeindex.ts- Application entry pointdata/- Data access and transformationschemas/- Zod schemas for data validationtools/- Utility tools and scriptssupport/- Supporting infrastructure (logging, validation, etc.)
Documentation
docs/- Project documentationarchitecture/- Architecture diagrams and descriptionsanalysis/- Data analysis and insights
Available Tools
meteoswissWeatherReport
Get official MeteoSwiss weather reports for Swiss regions.
Parameters:
region: Swiss region (north,south, orwest)north: Northern Switzerland (Zurich, Basel, Bern, Swiss Plateau)south: Southern Switzerland (Ticino and southern valleys)west: Western Switzerland/Romandy (Geneva, Lausanne, western Alps)
language: Report language (de,fr, orit- English not available)- Default:
de
- Default:
Returns: Detailed weather report with daily forecasts, temperatures, and regional conditions.
search
Search MeteoSwiss website content with pagination and multi-language support.
fetch
Retrieve full content from MeteoSwiss pages in various formats (markdown, text, HTML).
See the API documentation for detailed tool specifications.
Available Prompts
Pre-configured prompts for common weather queries:
German
wetterNordschweiz: Current weather report for Northern SwitzerlandwetterSchweiz: Interactive prompt for any Swiss region
French
meteoSuisseRomande: Current weather report for Western Switzerland (Romandy)
Italian
meteoTicino: Current weather report for Southern Switzerland (Ticino)
Debugging
For issues with Claude Desktop connections or other debugging needs, see our Debugging Guide.
Running Your Own Instance
Using Node.js
# Clone the repository
git clone https://github.com/eins78/mcp-server-meteoswiss.git
cd mcp-server-meteoswiss
# Install dependencies
pnpm install
# Start the server
pnpm start
# The server will be available at http://localhost:3000
Using Docker
# Run the latest version
docker run -p 3000:3000 -e USE_TEST_FIXTURES=false meteoswiss-mcp-server
# Or build your own
docker build -t my-meteoswiss-server .
docker run -p 3000:3000 my-meteoswiss-server
# Run with custom external port mapping
# Internal port 3000 mapped to external port 8080
docker run -p 8080:3000 -e PUBLIC_URL=http://localhost:8080 my-meteoswiss-server
# Run with custom hostname
docker run -p 80:3000 -e PUBLIC_URL=http://meteoswiss.example.com my-meteoswiss-server
Environment Variables
PORT- Server port (default: 3000)PUBLIC_URL- Full public URL including protocol and port for URL generationUSE_TEST_FIXTURES- Use test data instead of live API (default: false)DEBUG_MCHMCP- Enable debug logging (default: false)BIND_ADDRESS- Interface to bind to (default: 0.0.0.0)MAX_SESSIONS- Maximum concurrent sessions (default: 100)SESSION_TIMEOUT_MS- Session timeout in milliseconds (default: 300000)
Docker Port Mapping
When running in Docker with port mapping (e.g., -p 8080:3000), use the PUBLIC_URL environment variable to ensure URLs reflect the external port:
# Server listens on port 3000 internally, but is accessible on port 8080 externally
docker run -p 8080:3000 -e PORT=3000 -e PUBLIC_URL=http://localhost:8080 my-server
# For production with a domain name
docker run -p 443:3000 -e PORT=3000 -e PUBLIC_URL=https://api.example.com my-server
Local MCP Configuration
To use your local instance with Claude Desktop, add this to your configuration:
{
"mcpServers": {
"meteoswiss-local": {
"command": "npx",
"args": ["mcp-remote", "http://localhost:3000/mcp"]
}
}
}
Contributing
Contributions are welcome! Please ensure you follow the architecture guidelines outlined in the documentation.
License
ISC
Acknowledgments
- MeteoSwiss for providing the weather data
- Anthropic for the MCP specification
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。