MCP ExplorerExplorer

Mcp Python Demo

@chenchen0611on 10 months ago
2 MIT
FreeCommunity
AI Systems
This is a project based on MCP (Model Context Protocol) that integrates Zhiyu AI and Tencent Maps API, providing functionalities such as weather queries, geocoding, and web search.

Overview

What is Mcp Python Demo

mcp-python-demo is a project based on the Model Context Protocol (MCP) that integrates Zhipu AI and Tencent Maps API, providing functionalities such as weather queries, geocoding, and web search.

Use cases

Use cases include querying current weather conditions for cities in China, obtaining administrative codes for regions, and performing web searches based on user queries.

How to use

To use mcp-python-demo, clone the repository, install the dependencies using uv, configure the necessary environment variables in a .env file, and run either the web client or command line interface as per your preference.

Key features

Key features include a Zhipu AI-powered conversational system, Tencent Maps weather queries, geocoding services, intelligent web search, dual support for command line and web interfaces, and external network access.

Where to use

mcp-python-demo can be used in various fields such as weather forecasting, geographic information systems (GIS), and AI-driven applications that require location-based services.

Content

MCP Python Demo

这是一个基于 MCP (Model Context Protocol) 的项目,集成了智谱 AI 和腾讯地图 API,提供了天气查询、地理编码和网络搜索等功能。项目包含命令行界面和 Web 界面两种交互方式。

功能特性

  • 🤖 智谱 AI 驱动的对话系统
  • 🌤️ 腾讯地图天气查询
  • 🗺️ 地理编码查询
  • 🔍 智能网络搜索
  • 💻 命令行和 Web 界面双重支持
  • 🌐 支持外网访问

项目结构

mcp-python-demo/
├── src/
│   └── mcp_python_demo/
│       ├── __init__.py
│       ├── server.py
│       ├── client.py
│       ├── web_client.py
│       └── main.py
├── images/
├── pyproject.toml
├── .env
├── .env.example
├── .gitignore
└── README.md

快速开始

环境要求

  • Python >= 3.10
  • uv (推荐) 或 pip

安装

  1. 克隆项目:
git clone https://github.com/chenchen0611/mcp-python-demo.git
cd mcp-python-demo
  1. 使用 uv 安装依赖:
uv venv
source .venv/bin/activate  # Linux/Mac
# 或
.venv\Scripts\activate  # Windows
uv pip install -e .

配置

  1. 创建 .env 文件并配置必要的环境变量:
# 智谱AI配置
ZHIPU_API_KEY="your-zhipu-api-key"
ZHIPU_MODEL="glm-4-plus"
ZHIPU_BASE_URL="https://open.bigmodel.cn/api/paas/v4/"

# 腾讯地图API配置
TENCENT_MAP_API_KEY="your-tencent-map-api-key"
TENCENT_MAP_API_BASE="https://apis.map.qq.com/ws/"

运行

Web 界面(推荐)

cd src/mcp_python_demo
streamlit run web_client.py

访问以下地址之一:

  • 本地访问:http://localhost:8501
  • 局域网访问:http://[your-local-ip]:8501
  • 外网访问:http://[your-public-ip]:8501
页面配置

命令行界面

cd src/mcp_python_demo
# 使用本地服务器
uv run client.py --agent server.py

# 或使用远程服务器
uv run server.py
uv run client.py --agent http://127.0.0.1:8000

config

可用工具

1. 天气查询

async def query_weather(adcode: str, search_type: str = "now") -> str:
    """获取中国城市天气信息"""

2. 地理编码查询

async def query_adcode(region_name: str) -> str:
    """获取行政区划代码"""

3. 网络搜索

async def web_search(search_query: str, search_engine: str = "search_std") -> str:
    """使用智谱AI进行网络搜索"""

开发说明

添加新工具

  1. server.py 中使用 @mcp.tool() 装饰器添加新工具:
@mcp.tool()
async def your_tool(param1: str, param2: str = "default") -> str:
    """工具描述"""
    # 实现逻辑
    return result
  1. 工具会自动在客户端可用

环境变量

  • 所有敏感信息和配置都应该放在 .env 文件中
  • 不要将 .env 文件提交到版本控制系统
  • 参考 .env.example 进行配置

常见问题

  1. 连接错误

    • 检查服务器地址是否正确
    • 确认所有环境变量都已正确配置
    • 确保服务器正在运行且端口可访问
  2. API 调用失败

    • 验证 API 密钥是否有效
    • 检查网络连接
    • 确认 API 调用频率是否超限
  3. Web UI 无法访问

    • 检查防火墙设置
    • 确认端口是否开放
    • 验证服务器 IP 地址是否正确

致谢

本项目基于以下开源项目:

Tools

No tools

Comments

Recommend MCP Servers

View All MCP Servers