mcp-webresearch
by: mzxrai
MCP web research server (give Claude real-time info from the web)
📌Overview
Purpose: To facilitate real-time web research for the Claude AI by integrating a Model Context Protocol (MCP) server that captures and processes web content.
Overview: The MCP Web Research Server enables users to conduct in-depth research by seamlessly incorporating online information into the Claude AI system. It enhances the AI's capabilities by allowing it to extract content from the web and manage research sessions effectively.
Key Features:
-
Google search integration: Enables the AI to perform searches and retrieve relevant web results in real-time for better-informed responses.
-
Webpage content extraction: Allows the AI to visit specific URLs and extract content from those pages, enriching the data available for analysis and responses.
-
Research session tracking: Maintains a log of visited pages, search queries, and captured content, helping users keep track of their research activities.
-
Screenshot capture: Provides the ability to take and store screenshots of webpages, facilitating easy reference and citation of sources during research.
MCP Web Research Server
A Model Context Protocol (MCP) server for web research that brings real-time information into Claude and enables easy research on any topic.
Features
- Google search integration
- Webpage content extraction
- Research session tracking (visited pages, search queries, etc.)
- Screenshot capture
Prerequisites
- Node.js >= 18 (includes
npm
andnpx
) - Claude Desktop app
Installation
Ensure you have installed the Claude Desktop app and have npm installed.
Add the following entry to your claude_desktop_config.json
(on Mac: ~/Library/Application Support/Claude/claude_desktop_config.json
):
{
"mcpServers": {
"webresearch": {
"command": "npx",
"args": ["-y", "@mzxrai/mcp-webresearch@latest"]
}
}
}
This configuration allows Claude Desktop to automatically start the web research MCP server when needed.
Usage
Start a chat with Claude and send a prompt that benefits from web research. You can use the agentic-research
prompt provided through this package for deeper web research. Access it in Claude Desktop by clicking the Paperclip icon in the chat input, then select Choose an integration → webresearch → agentic-research.
Tools
-
search_google
Performs Google searches and extracts results
Arguments:{ query: string }
-
visit_page
Visits a webpage and extracts its content
Arguments:{ url: string, takeScreenshot?: boolean }
-
take_screenshot
Takes a screenshot of the current page
No arguments required
Prompts
agentic-research
A guided research prompt that instructs Claude to:
- Start with broad searches to understand the topic landscape
- Prioritize high-quality, authoritative sources
- Iteratively refine the research direction based on findings
- Keep you informed and allow interactive guidance
- Always cite sources with URLs
Resources
Two MCP resources are exposed: captured webpage screenshots and the research session.
Screenshots
Screenshots taken are saved as MCP resources accessible in Claude Desktop via the Paperclip icon.
Research Session
The server maintains a research session including:
- Search queries
- Visited pages
- Extracted content
- Screenshots
- Timestamps
Suggestions
For better results, if not using the agentic-research
prompt, suggest high-quality sources to Claude. For example, prompt "news today from Reuters or AP" instead of just "news today".
Problems
This is pre-alpha software and uses AIGC technology, so expect bugs.
If issues arise, check Claude Desktop's MCP logs:
tail -n 20 -f ~/Library/Logs/Claude/mcp*.log
Development
# Install dependencies
pnpm install
# Build the project
pnpm build
# Watch for changes
pnpm watch
# Run in development mode
pnpm dev
Requirements
- Node.js >= 18
- Playwright (installed automatically as a dependency)
Verified Platforms
- macOS
- Linux (not yet verified)
License
MIT
Author
mzxrai