Duckduckgo MCP Server

🚀 DuckDuckGo Search MCP Server

An MCP (Model Context Protocol) server that provides DuckDuckGo web search functionality, with additional features for content scraping and parsing.

🚀 Quick Start

This is a Model Context Protocol (MCP) server that offers DuckDuckGo web search capabilities, along with content scraping and parsing features.

✨ Features

• Web Search: Conduct advanced rate-limited searches on DuckDuckGo with formatted results.
• Content Scraping: Fetch and parse web page content for intelligent text extraction.
• Rate Limiting: Built - in rate protection for both searches and content scraping.
• Error Handling: Comprehensive error capture and logging.
• LLM - Friendly Output: Formatted results designed specifically for large language models.

📦 Installation

Install via Smithery

To automatically install the DuckDuckGo search server via Smithery:

npx -y @smithery/cli install @nickclyde/duckduckgo-mcp-server --client claude

Install via uv

Install directly from PyPI using uv:

uv pip install duckduckgo-mcp-server

💻 Usage Examples

Run on Claude Desktop

1. Download Claude Desktop.
2. Create or edit the Claude Desktop configuration:
   • On macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
   • On Windows: %APPDATA%\Claude\claude_desktop_config.json Add the following configuration:

{
    "mcpServers": {
        "ddg-search": {
            "command": "uvx",
            "args": ["duckduckgo-mcp-server"]
        }
    }
}

3. Restart Claude Desktop.

Development

For local development, you can use the MCP CLI:

# Run with MCP Inspector
mcp dev server.py

# Install to Claude Desktop for testing
mcp install server.py

Basic Usage

Here are the available tools and their usage:

1. Search Tool

async def search(query: str, max_results: int = 10) -> str

Perform a web search on DuckDuckGo and return formatted results.

Parameters:

• query: The search query string.
• max_results: The maximum number of results to return (default is 10).

Returns: A formatted string containing titles, URLs, and excerpts.

2. Content Scraping Tool

async def fetch_content(url: str) -> str

Scrape and parse content from a web page.

Parameters:

• url: The web page address to scrape content from.

Returns: Cleaned and formatted text content.

🔧 Technical Details

Rate Limiting

• Search: Maximum of 30 requests per minute.
• Content Scraping: Maximum of 20 requests per minute.
• Automatic queue management and waiting time.

Result Processing

• Remove ads and irrelevant content.
• Clean up DuckDuckGo redirect URLs.
• Format results to optimize the experience for large language models.
• Appropriately truncate long content.

Error Handling

• Comprehensive error capture and reporting.
• Detailed logging through the MCP context.
• Graceful degradation in case of rate limits or timeouts.

🤝 Contribution

Issues and pull requests are welcome! Some areas for improvement include:

• Documentation: Enhance existing documentation by adding more explanations and tutorials.
• Code Quality: Review and optimize the codebase to improve code quality.
• Feature Expansion: Explore new features and integrate them into the project.

Visit the GitHub repository for contribution guidelines and other details.

📄 License

The license information for this project can be found in the LICENSE file.