Github 9a2

🚀 Tavily-MCP Project Documentation

Tavily-MCP is a powerful tool that integrates with the Claude desktop application, enabling enhanced search and content extraction capabilities. This documentation provides a comprehensive guide on installation, configuration, usage, and troubleshooting.

🚀 Quick Start

Get up and running with Tavily-MCP in no time using the following installation and configuration steps.

📦 Installation

1. Install via npm

Step 1: Install Dependencies

Run the following command in your terminal or command prompt to install Tavily-MCP:

npm install tavily-mcp

Step 2: Start the Server

After successful installation, start the Tavily-MCP server:

tavily-mcp start

2. Manual Installation

Step 1: Clone the Repository

Run the following commands in your terminal or command prompt to clone the Tavily-MCP repository:

git clone https://github.com/tavily-ai/tavily-mcp.git
cd tavily-mcp

Step 2: Install Dependencies

Once inside the repository directory, install all dependencies:

npm install

Step 3: Build the Project

Build the project to generate an executable file:

npm run build

📚 Configuration Guide

Configure in the Claude Desktop Application

1. Open the configuration file:

   • macOS:

     code ~/Library/Application\ Support/Claude/claude_desktop_config.json
     

   • Windows:

     code %APPDATA%\Claude\claude_desktop_config.json
     

2. Add the Tavily-MCP server configuration:

   {
     "mcpServers": {
       "tavily-mcp": {
         "command": "npx",
         "args": ["-y", "tavily-mcp@1.0.0"],
         "env": {
           "TAVILY_API_KEY": "your-api-key-here"
         }
       }
     }
   }
   

   • Replace your-api-key-here with your actual Tavily API key.

💻 Usage Guide

Use in the Claude Desktop Application

1. After installation, completely close and reopen the Claude desktop application for the configuration to take effect.
2. A hammer icon will appear at the bottom left of the application, indicating that MCP tools are available. Click the icon to view more detailed information.

💻 Usage Examples

Tavily Search Examples

1. General Web Search:

   Can you find the latest developments in the field of quantum computing?
   

2. News Search:

   Find news articles about AI startups from the past week.
   

3. Domain-Specific Search:

   Search for research on climate change on nature.com and sciencedirect.com.
   

Tavily Extract Example

1. Extract Article Content:

   Extract the main content from this article https://example.com/article.
   

Combined Use of Search and Extract

You can combine the Tavily-Search and Tavily-Extract tools to perform more complex tasks.

Search for news articles about AI startups from the past week and extract the main content of each article to generate a detailed report.

🔧 Troubleshooting

Common Issues

1. Server Not Found

   • Verify the npm installation status by running npm --version.
   • Check the syntax of the Claude desktop application configuration file by running code ~/Library/Application\ Support/Claude/claude_desktop_config.json.
   • Ensure that Node.js is correctly installed by running node --version.

2. NPX-Related Issues

   • If you encounter errors related to npx, replace "command": "npx" in the configuration with the full path of npx.

     which npx
     

     Then replace it in the configuration with:

     "command": "/full/path/to/npx"
     

3. API Key Issues

   • Confirm that your Tavily API key is valid.
   • Check that the key is correctly set in the configuration file.
   • Ensure that the key does not contain spaces or quotes.

📄 Acknowledgments

• Model Context Protocol for providing the MCP specification.
• Anthropic for providing the Claude desktop application.