Web Search MCP Server (Standalone)

A Model Context Protocol (MCP) server that provides web search capabilities using multiple search engines (DuckDuckGo, Bing, and Startpage) with parallel execution and result deduplication.

Features

• Multi-engine search: Searches DuckDuckGo, Bing, and Startpage simultaneously
• Parallel execution: All searches run concurrently for faster results
• Result deduplication: Removes duplicate URLs from combined results
• Source attribution: Shows which search engine provided each result
• Comprehensive logging: Logs all operations to web-search.log
• Standalone: Pre-configured virtual environment included - no setup required
• Two tools available:
  • SearchWeb: Multi-engine web search
  • FetchPageContent: Extract text content from web pages

Installation

Option 1: Standalone (Pre-configured venv included)

1. Extract the files:

   unzip web-search-mcp.zip -d ~/.mcp/
   

2. Make start script executable:

   chmod +x ~/.mcp/web-search/start.sh
   

3. Add to Q CLI MCP configuration:

   q mcp add web-search ~/.mcp/web-search/start.sh
   

4. Verify installation and test:

   q mcp status --name web-search
   q chat
   # In chat: "search web for python libraries"
   

Option 2: Manual Setup (If venv is missing)

1. Navigate to the directory:

   cd ~/.mcp/web-search
   

2. Create virtual environment:

   python3 -m venv venv
   

3. Activate virtual environment and install dependencies:

   source venv/bin/activate
   pip install -r requirements.txt
   deactivate
   

4. Make start script executable:

   chmod +x start.sh
   

5. Add to Q CLI MCP configuration:

   q mcp add web-search ~/.mcp/web-search/start.sh
   

6. Test the installation:

   q mcp status --name web-search
   

Files Included

• server.py - Main MCP server code
• start.sh - Startup script
• requirements.txt - Python dependencies
• venv/ - Pre-configured virtual environment with all dependencies
• README.md - This documentation

Usage

Start a new Q CLI chat session. The server provides two tools:

SearchWeb

Search query: "python web scraping libraries"

FetchPageContent

Fetch content from: "https://example.com"

Logs

All operations are logged to web-search.log in the server directory with timestamps and detailed information about search requests and results.

Troubleshooting

• Permission denied: Ensure start.sh is executable (chmod +x start.sh)
• No search results: Check the log file for error messages
• Debugging: Check web-search.log for detailed debugging information

Requirements

• Q CLI with MCP support
• Internet connection for web searches
• macOS/Linux (Python 3.11+ included in venv)