MCP2Tavily
A MCP protocol server that implements web search functionality using the Tavily API.
Prerequisites
- Python 3.11+
- UV package manager
- Tavily API key
Installation
- Clone the repository
git clone <repository-url>
cd mcp2tavily
- Create and edit the
.envfile
# Create .env file
touch .env
# Add your Tavily API key to .env
echo "TAVILY_API_KEY=your_api_key_here" > .env
- Set up virtual environment with UV
# Create and activate virtual environment
uv venv
source .venv/bin/activate # On Windows use: .venvScriptsactivate
- Install dependencies
uv sync
Usage
Install as Claude extension
fastmcp install mcp2tavily.py
Development mode with MCP Inspector
To test the functionality using MCP Inspector:
fastmcp dev mcp2tavily.py
Once running, you can access the MCP Inspector at: http://localhost:5173
Available Tools
search_web(query: str): Search the web using Tavily APIsearch_web_info(query: str): Same as above, with Chinese description
Environment Variables
TAVILY_API_KEY: Your Tavily API key (required)
Step-by-Step Guide
手动添加Cline Continue Claude
Cline Continue Claude的MCP JSON FILE
"mcp2tavily": {
"command": "uv",
"args": [
"run",
"--with",
"fastmcp",
"--with",
"python-dotenv",
"--with",
"tavily-python",
"fastmcp",
"run",
"C:\Users\你的真实路径\mcp2tavily.py"
],
"env": {
"TAVILY_API_KEY": "API密钥"
}
}
Cline

Cline

Cline

EXAMPLE

Tavily Web Search
Project Details
- mcp2everything/mcp2tavily
- Last Updated: 2/14/2025
Recomended MCP Servers
🤖 A TypeScript implementation of a Sentry MCP (Modern Context Protocol) tool that allows AI agents to access...
Stata MCP Extension for VS Code and Cursor IDE
MCP server to extract contents from a PDF file
A Model Context Protocol (MCP) server for Windows desktop automation using AutoIt.
A Model Context Protocol (MCP) service for managing Django migrations in distributed environments.
A Model Context Protocol (MCP) server implementation that provides EMQX MQTT broker interaction.
302 BrowserUse MCP
Zero Trust Agentic Access based MCP Server Reference Implementation
Simple MCP Server Implementation
MCP for judged outputs between two AIs for a better output





