MCP Server
An extensible MCP server with weather and time tools for Claude Desktop. It provides current time with automatic timezone detection, weather alerts for US states, and 5-period forecasts for coordinates.
README
MCP Server
An extensible Model Context Protocol (MCP) server with weather and time tools. Built for learning — connect it to Claude Desktop and start chatting.
Getting Started
1. Clone & Install
git clone <repo-url> && cd mcp-server
python3 -m venv .venv
source .venv/bin/activate # macOS / Linux / WSL
pip install -e .
Or with uv:
git clone <repo-url> && cd mcp-server
uv venv
source .venv/bin/activate
uv pip install -e .
2. Connect to Claude Desktop
Run the setup script to automatically write the Claude Desktop config file:
macOS:
python setup_claude_agent.py
This writes ~/Library/Application Support/Claude/claude_desktop_config.json pointing to your project's .venv/bin/python.
If your project is in a non-standard location, pass it explicitly:
python setup_claude_agent.py --project-path /path/to/mcp-server
Windows (via WSL):
python3 setup_claude_agent.py --windows --win-user <YourWindowsUser> --wsl-user <YourWSLUser>
For example:
python3 setup_claude_agent.py --windows --win-user Cam --wsl-user cam
This writes %APPDATA%\Claude\claude_desktop_config.json and configures Claude Desktop to launch the server through WSL.
3. Restart Claude Desktop
Close and reopen Claude Desktop. To verify the server is connected, go to Settings → Developer — you should see mcp-server with a green running badge.
4. Try It Out
Ask Claude any of the following:
- "What time is it?" — uses the
get_current_timetool - "Get weather alerts for California" — uses the
weather://alerts/CAresource - "What's the forecast for latitude 40.7128, longitude -74.0060?" — uses the forecast resource
You don't need to start the server manually — Claude Desktop launches it automatically.
Available Tools & Resources
| Type | Name | What it does |
|---|---|---|
| Tool | get_current_time |
Returns current time with automatic timezone detection |
| Resource | weather://alerts/{state} |
Weather alerts for a US state (e.g. CA, NY) |
| Resource | weather://forecast/{lat}/{lon} |
5-period forecast for coordinates |
| Prompt | analyze_weather_prompt |
Guides Claude through a full weather analysis |
| Prompt | timezone_helper_prompt |
Helps with timezone checks, conversions, comparisons |
Adding Your Own Tools
Open src/mcp_server/server.py and add a function with the @mcp.tool() decorator:
@mcp.tool()
async def my_tool(param: str) -> str:
"""Description of what this tool does."""
return f"Result for {param}"
Restart Claude Desktop to pick up the change.
For resources and prompts, see CONTRIBUTING.md.
Running Tests
pip install -e ".[dev]" # or: uv pip install -e ".[dev]"
pytest
Project Structure
src/mcp_server/
├── server.py # MCP server entry point — register tools here
└── tools/
├── weather/ # Weather alerts & forecasts (NWS API)
├── time/ # IP-based timezone & current time
└── conversation/ # Conversation tools
License
MIT
Resources
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。
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。
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。
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。
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。
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。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
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