MCP DuckDuckResearch

What is MCP DuckDuckResearch?

MCP DuckDuckResearch is an MCP server that combines DuckDuckGo search capabilities with web page content extraction and screenshot functionality. It bridges searching for information and accessing web content programmatically.

How to use MCP DuckDuckResearch?

Install via git clone, npm install, and npm run build. Configure the MCP server in Cline’s MCP settings file by adding a JSON entry with "command": "node" and "args": ["path/to/mcp-duckduckresearch/build/index.js"]. Use tools such as search_duckduckgo, visit_page, and take_screenshot in Roo Code via <use_mcp_tool> blocks.

Key features of MCP DuckDuckResearch

• DuckDuckGo search with configurable safe search and region
• Web page content extraction and conversion to Markdown
• Screenshot capture with automatic size optimization
• Robust error handling and bot detection protection
• Browser initialization and cleanup managed automatically

Use cases of MCP DuckDuckResearch

• Searching the web and retrieving full-text content from result pages
• Taking screenshots of web pages for documentation or analysis
• Integrating web search and page scraping into AI agent workflows
• Automating research tasks that require both search and content extraction