Google Search MCP Server
Enables Google search functionality through the Google Custom Search API with enterprise-grade security authentication. Supports streamable HTTP transport protocol for real-time search operations with configurable result limits.
README
Google Search MCP Server
一个基于FastMCP框架的Google搜索服务器,提供Google自定义搜索功能。支持Streamable HTTP传输协议,具备企业级安全认证机制。
✨ 主要特性
- 🔍 Google搜索集成: 使用Google Custom Search API提供强大的搜索功能
- 🚀 Streamable HTTP: 支持实时通信的高性能HTTP传输协议
- 🔐 企业级安全: 内置Bearer Token认证机制,支持生产环境安全部署
- ⚙️ 灵活配置: 通过环境变量轻松配置服务器参数
- 📦 现代化依赖管理: 支持UV和pip两种依赖管理方式
- 🛠️ 开发友好: 完整的开发工具链,包括测试、格式化、类型检查
- 📚 完整文档: 详细的服务器和客户端配置指南
🚀 快速开始
环境要求
- Python 3.8+
- Google API Key
- Google Custom Search Engine ID
安装
使用UV(推荐)
# 安装UV
powershell -c "irm https://astral.sh/uv/install.sh | iex"
# 克隆项目
git clone https://github.com/yourusername/google-search-mcp.git
cd google-search-mcp
# 创建虚拟环境
uv venv
.venv\Scripts\activate
# 安装依赖
uv pip install -e .
使用pip
# 克隆项目
git clone https://github.com/yourusername/google-search-mcp.git
cd google-search-mcp
# 创建虚拟环境
python -m venv venv
venv\Scripts\activate
# 安装依赖
pip install -r requirements.txt
配置
- 复制环境变量模板文件:
copy .env.example .env
- 编辑
.env文件,配置必要的环境变量:
# Google API配置
GOOGLE_API_KEY=your_google_api_key_here
GOOGLE_CSE_ID=your_custom_search_engine_id_here
# 服务器配置
HTTP_HOST=127.0.0.1
HTTP_PORT=9000
# 认证配置(生产环境建议启用)
ENABLE_AUTH=true
API_TOKEN=your_secure_api_token_here
运行服务器
# 启动服务器
python main.py
服务器将在 http://127.0.0.1:9000/mcp/ 启动。
🔧 配置选项
环境变量
| 变量名 | 描述 | 默认值 | 必需 |
|---|---|---|---|
GOOGLE_API_KEY |
Google API密钥 | - | ✅ |
GOOGLE_CSE_ID |
Google自定义搜索引擎ID | - | ✅ |
HTTP_HOST |
服务器主机地址 | 127.0.0.1 |
❌ |
HTTP_PORT |
服务器端口 | 9000 |
❌ |
ENABLE_AUTH |
是否启用认证 | false |
❌ |
API_TOKEN |
API认证令牌 | - | 认证启用时必需 |
认证配置
开发环境(无认证)
ENABLE_AUTH=false
生产环境(启用认证)
ENABLE_AUTH=true
API_TOKEN=your_secure_api_token_here
📖 使用方法
服务器API
服务器提供一个搜索工具:
search_google
执行Google搜索并返回格式化结果。
参数:
query(str): 搜索查询字符串num_results(int, 可选): 返回结果数量,默认为5auth_token(str, 可选): 认证令牌(启用认证时必需)
返回值:
{
"success": true,
"results": [
{
"title": "结果标题",
"link": "结果链接",
"snippet": "结果摘要"
}
],
"total_results": "1000000"
}
客户端连接
基本连接(无认证)
import asyncio
from fastmcp import Client
async def main():
config = {
"mcpServers": {
"google-search": {
"url": "http://127.0.0.1:9000/mcp/",
"transport": "streamable-http"
}
}
}
client = Client(config)
async with client:
result = await client.call_tool(
"search_google",
{"query": "Python编程", "num_results": 3}
)
print(result)
if __name__ == "__main__":
asyncio.run(main())
带认证的连接
import asyncio
from fastmcp import Client
async def main():
config = {
"mcpServers": {
"google-search": {
"url": "http://127.0.0.1:9000/mcp/",
"transport": "streamable-http",
"headers": {
"Authorization": "Bearer your_api_token_here"
}
}
}
}
client = Client(config)
async with client:
result = await client.call_tool(
"search_google",
{"query": "人工智能", "num_results": 5}
)
print(result)
if __name__ == "__main__":
asyncio.run(main())
运行示例
项目提供了完整的客户端示例:
# 运行客户端示例
python client_example.py
# 使用命令行工具
python run_client.py --help
🛠️ 开发
安装开发依赖
# 使用UV
uv pip install -e ".[dev]"
# 使用pip
pip install -e ".[dev]"
代码格式化
# 格式化代码
uv run black .
uv run isort .
类型检查
# 运行类型检查
uv run mypy .
测试
# 运行测试
uv run pytest
# 运行测试并生成覆盖率报告
uv run pytest --cov=google_search_mcp
代码检查
# 运行代码检查
uv run flake8 .
📁 项目结构
google-search-mcp/
├── main.py # 主服务器文件
├── client_example.py # 客户端示例
├── run_client.py # 命令行客户端工具
├── client_config.json # 客户端配置文件
├── pyproject.toml # 项目配置文件
├── requirements.txt # pip依赖文件
├── uv.lock # UV锁定文件
├── .env.example # 环境变量模板
├── .env # 环境变量文件(git忽略)
├── README.md # 项目说明文档
├── README_CLIENT.md # 客户端配置指南
├── CLIENT_SETUP.md # 客户端设置详细文档
└── UV_USAGE.md # UV使用指南
🔍 详细文档
🤝 贡献
欢迎贡献代码!请遵循以下步骤:
- Fork本项目
- 创建功能分支 (
git checkout -b feature/AmazingFeature) - 提交更改 (
git commit -m 'Add some AmazingFeature') - 推送到分支 (
git push origin feature/AmazingFeature) - 创建Pull Request
📄 许可证
本项目采用MIT许可证 - 查看 LICENSE 文件了解详情。
🙏 致谢
- FastMCP - 提供MCP服务器框架
- Google Custom Search API - 提供搜索功能
- UV - 现代化的Python包管理器
📞 支持
如果您遇到问题或有建议,请:
注意: 使用本服务器需要有效的Google API密钥和Custom Search Engine ID。请确保遵守Google API的使用条款和服务限制。
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.