searxng-mcp-server

shgsousa/searxng-mcp-server

3.1

If you are the rightful owner of searxng-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

SearXNG MCP Server is a comprehensive Model Context Protocol server offering advanced web search, content scraping, and AI-powered summarization through a Gradio web interface.

Tools
3
Resources
0
Prompts
0

SearXNG MCP Server

A comprehensive Model Context Protocol (MCP) server that provides a powerful Gradio web interface for performing advanced web searches, content scraping, and AI-powered content summarization using SearXNG as the backend.

Features

🔍 Advanced Web Search

  • Multi-Engine Support: Google, Bing, Brave, DuckDuckGo, Yahoo, and more
  • Flexible Result Formats:
    • Summary: Fast results with titles, URLs, and snippets
    • Full Content: Complete webpage content extraction with intelligent filtering
    • AI Summary: AI-generated summaries of full content (requires OpenAI/OpenRouter API)
  • Advanced Filtering: Time range, language, SafeSearch, and result count controls
  • Custom SearXNG Instance: Override default instance with your own SearXNG deployment

🌐 Web Content Scraping

  • Intelligent Content Extraction: Automatically filters out navigation, ads, and boilerplate
  • Multi-Site Optimization: Special handling for Wikipedia, tech blogs, and various website types
  • Markdown Conversion: Clean, structured markdown output preserving links and formatting
  • AI Summarization: Optional AI-powered content summaries

🛠 Diagnostics & Utilities

  • Connection Testing: Comprehensive SearXNG instance diagnostics with troubleshooting tips
  • Date/Time Service: Current timestamp utility for logging and reference
  • Docker Support: Complete containerized setup with SearXNG included

🤖 MCP Integration

  • API Endpoints: All functions available via REST API for programmatic access
  • AI System Integration: MCP-enabled for seamless integration with AI assistants
  • Multi-Tab Interface: Organized web interface with dedicated sections for each feature

Setup and Installation

  1. Clone the repository:

    git clone https://github.com/shgsousa/searxng-mcp-server.git
cd searxng-mcp-server

  2. Install the required packages (choose one method):

    # Using pip with requirements.txt
pip install -r requirements.txt

# OR using the package setup
pip install -e .

  3. Run the server:

    python main.py

Usage

The server provides a comprehensive web interface with multiple tabs for different functionality:

🔍 Search Tab

  1. Open the Gradio interface in your web browser (default: http://localhost:7870)
  2. Navigate to the Search tab
  3. Enter your search query in the text box
  4. Select a search engine from the dropdown menu
  5. Choose the result format:
    • Summary: Fast results with basic information
    • Full: Complete webpage content for each result
    • AI Summary: AI-generated summaries (requires API key)
  6. Configure advanced options (time range, language, SafeSearch, max results)
  7. Optionally specify a custom SearXNG instance URL
  8. Click Submit to perform the search

🌐 Web Scraper Tab

  1. Navigate to the Web Scraper tab
  2. Enter the URL of the webpage you want to scrape
  3. Optionally enable AI Summarization for condensed content
  4. Click Submit to extract and process the content

🛠 Diagnostics Tab

  1. Navigate to the Diagnostics tab
  2. Optionally enter a custom SearXNG instance URL to test
  3. Click Submit to run comprehensive connectivity tests
  4. Review the detailed diagnostic report and troubleshooting tips

🕒 Date & Time Tab

  1. Navigate to the Date & Time tab
  2. Click Submit to get the current system date and time

🔌 API Access

All functionality is available via REST API endpoints:

  • /api/search - Web search functionality
  • /api/scrape - Web content scraping
  • /api/diagnostics - SearXNG diagnostics
  • /api/datetime - Current date/time

Configuration

The server supports configuration through environment variables or a .env file:

Core Configuration

By default, the MCP server uses a local SearXNG instance at http://localhost:8080. To use a different SearXNG instance:

Option 1: Environment Variable

# On Windows PowerShell
$env:SEARXNG_URL="https://your-searxng-instance.com"
python main.py

Option 2: .env File

Create a .env file in the project root (you can use .env.example as a template):

SEARXNG_URL=https://your-searxng-instance.com

Option 3: Direct Configuration

Edit the config.py file:

SEARXNG_URL = "https://your-searxng-instance.com"

AI Features Configuration

AI features (content summarization, enhanced processing) require OpenAI or OpenRouter API credentials:

OpenAI Configuration:

OPENAI_API_URL=https://api.openai.com/v1
OPENAI_API_TOKEN=your_openai_api_key_here
OPENAI_MODEL=gpt-4o-mini  # Optional, defaults to gpt-3.5-turbo

OpenRouter Configuration (Alternative):

OPENAI_API_URL=https://openrouter.ai/api/v1
OPENAI_API_TOKEN=your_openrouter_api_key_here
OPENAI_MODEL=openai/gpt-4o-mini  # Model format for OpenRouter

Advanced Configuration Options

# Server Configuration
HOST=0.0.0.0              # Gradio server host (default: 127.0.0.1)  
PORT=7870                 # Gradio server port (default: 7870)
MCP_ENABLED=true          # Enable MCP server functionality

# Search Configuration  
MAX_RESULTS=20            # Maximum search results per query
DEFAULT_RESULTS=10        # Default number of results
DEFAULT_ENGINE=google     # Default search engine

# Content Processing
CONTENT_TIMEOUT=30        # Timeout for webpage content fetching (seconds)
SUMMARY_MAX_LENGTH=10000  # Maximum content length for AI summarization

Running SearXNG Locally

The project includes a complete Docker Compose setup that runs both SearXNG and the MCP server:

docker-compose up -d

This will:

  1. Start a SearXNG instance at http://localhost:8080
  2. Start the SearXNG MCP server at http://localhost:7870
  3. Configure them to work together automatically

The docker-compose.yml file includes all necessary configuration and volume mappings.

MCP Integration

The server is MCP-enabled through Gradio's launch function, allowing it to be integrated with AI systems as a tool. When MCP_ENABLED=true is set in the configuration, the server provides:

  • Standardized Tool Interface: Compatible with MCP-enabled AI assistants
  • Function Descriptions: Comprehensive documentation for each available function
  • Type Safety: Proper parameter validation and error handling
  • Multi-Modal Support: Text, JSON, and markdown output formats

Dependencies

Core Dependencies

  • gradio: Web interface framework with MCP integration support
  • requests: HTTP client for API calls and web scraping
  • beautifulsoup4: HTML parsing and content extraction
  • html2text: Clean HTML to Markdown conversion
  • python-dotenv: Environment variable management
  • openai: AI integration for content summarization (optional)

Development Dependencies

  • markdown: Markdown processing and validation
  • pytest: Testing framework (for future test development)
  • black: Code formatting
  • flake8: Code linting

Optional Dependencies

  • docker: For containerized deployment
  • antml-mcp: Enhanced MCP integration features

Development

Running with VS Code Tasks

This project includes VS Code tasks for easy execution:

  1. Press Ctrl+Shift+P and select "Tasks: Run Task"
  2. Select "Run SearXNG MCP Server"

Docker Support

The project includes Docker Compose support for easy deployment:

Start all services:

docker-compose up -d

Stop all services:

docker-compose down

Rebuild after changes:

docker-compose up -d --build

View logs:

docker-compose logs -f

Local Development Setup

  1. Clone the repository:

    git clone https://github.com/shgsousa/searxng-mcp-server.git
cd searxng-mcp-server

  2. Install dependencies:

    # Using pip with requirements.txt
pip install -r requirements.txt

# OR using the package setup
pip install -e .

  3. Configure environment variables:

    # Copy the example environment file
cp .env.example .env

# Edit .env with your preferred settings

  4. Run the server:

    python main.py

API Documentation

When the server is running, API documentation is automatically available:

  • Interactive Documentation: Visit /docs endpoint for Swagger UI
  • API Schema: Available at /api/docs for programmatic access
  • OpenAPI Specification: JSON schema available at /openapi.json

Function Documentation

All major functions include comprehensive docstrings with:

  • Parameter descriptions with types and constraints
  • Return value specifications with example structures
  • Usage examples for common scenarios
  • Error handling information and troubleshooting tips
  • Performance considerations and best practices

Troubleshooting

Common Issues

SearXNG Connection Issues:

  • Use the Diagnostics tab to test your SearXNG instance
  • Ensure SearXNG is running and accessible at the configured URL
  • Check firewall settings if using custom instances

AI Summarization Not Working:

  • Verify OPENAI_API_TOKEN is set in your environment
  • Check OPENAI_API_URL points to the correct endpoint
  • Ensure you have sufficient API credits/quota

Content Scraping Issues:

  • Some websites block automated requests
  • Try different user agents or request headers
  • Respect robots.txt and website terms of service

Docker Issues:

  • Ensure Docker and Docker Compose are installed
  • Check that ports 7870 and 8080 are available
  • Review container logs with docker-compose logs

License

Contributing

Pull requests are welcome! For major changes, please open an issue first to discuss what you would like to change.

Contributing Guidelines

  1. Fork the repository and create a feature branch
  2. Add tests for new functionality where applicable
  3. Update documentation including docstrings and README
  4. Follow code style guidelines (use black for formatting)
  5. Test thoroughly with different SearXNG instances and configurations
  6. Submit a pull request with a clear description of changes

Development Roadmap

  • Enhanced content filtering and extraction algorithms
  • Support for additional AI providers (Anthropic, Cohere, etc.)
  • Advanced search result ranking and relevance scoring
  • Built-in caching mechanisms for improved performance
  • WebSocket support for real-time search updates
  • Plugin system for custom content processors