MCP Tavily Proxy

🚀 Tavily MCP Server with Proxy Support

An AI-driven network search model context protocol (MCP) server that uses Tavily's search API and adds support for HTTP/HTTPS proxy configuration. This server allows large language models (LLMs) to perform complex network searches, directly answer questions, search for recent news articles, and extract relevant content through proxy servers.

🚀 Quick Start

The Tavily MCP Server with Proxy Support is an AI - powered network search server. It utilizes Tavily's search API and adds proxy configuration capabilities, enabling large language models to perform network searches, answer questions, and extract news content.

✨ Features

• All original features of the Tavily MCP Server.
• Support for HTTP/HTTPS proxies via environment variables.
• Enhanced logging for proxy configuration.
• Robust error handling for proxy - related issues.

💻 Usage Examples

Basic Usage

import mcp_server_tavily as tavily

# Create an MCP server instance with preset context and proxy configuration
server = tavily.createServer()

# Initialize the server with custom context
context = {
    "temperature": 0.7,
    "max_tokens": 2048,
    "model_name": "gpt-3.5-turbo",
}
server = tavily.createServer(context)

# Process a query
response = server.process_query("Tell me about artificial intelligence")
print(response)

📦 Installation

Using pip

pip install mcp-tavily-proxy

Or if you have uv installed:

uv pip install mcp-tavily-proxy

Building from Source

Clone this repository, build, and then install the program:

git clone https://github.com/tulong66/mcp-tavily-proxy.git
cd mcp-tavily-proxy
uv build
uv pip install .

📚 Documentation

API Key and Proxy Settings

The server requires a Tavily API key and supports proxy configuration via environment variables:

1. Set the required environment variables:

# Tavily API Key
export TAVILY_API_KEY=your_api_key_here

# Proxy settings (if needed)
export HTTP_PROXY=http://your-proxy:port
export HTTPS_PROXY=http://your-proxy:port

2. Or provide the API key via command - line arguments:

python -m mcp_server_tavily --api-key=your_api_key_here

Configuration as a Cloudflare Worker

To use this library as a Cloudflare Worker, refer to CLOUDFLARE.md.

Available Tools

Debugging

# Enable detailed logging for debugging
export MCP_DEBUG=true

Contribution Guidelines

If you want to contribute, refer to CONTRIBUTING.md.

📄 License

This project is licensed under the MIT License. See LICENSE for details.