Web Scout

Web-Scout: AI-Powered Search with LLM Summarization

A FastAPI application that performs web searches using DuckDuckGo and generates AI-powered summaries using Google's Gemini AI.

- Web search using DuckDuckGo
- AI summarization using Gemini 2.5 Flash
- Two output modes: Summary and Detailed analysis
- Docker & Docker Compose ready
- Secure API key management via environment variables
- MCP (Model Context Protocol) support over HTTP

- Docker and Docker Compose installed
- Google Gemini API key
-

Clone the repository(or navigate to your project directory)

Add your Gemini API keyto the.envfile:

GEMINI_API_KEY=your_actual_gemini_api_key_here

The application will be available athttp://localhost:8000

curl "http://localhost:8000/search?query=artificial+intelligence" # or explicitly specify mode=summary curl "http://localhost:8000/search?query=artificial+intelligence&mode=summary"

curl "http://localhost:8000/search?query=artificial+intelligence&mode=detailed"

{ "query": "your search query", "mode": "summary", "summary": "AI-generated analysis...", "sources_used": 10 }

Web-Scout can also function as an MCP (Model Context Protocol) server, allowing AI assistants to perform web searches directly through tools.

- Web Search Tool: Perform web searches with AI summarization
- Dual Mode Support: Both summary and detailed analysis modes
- HTTP Transport: MCP over HTTP protocol for client integration
- JSON Responses: Structured output for easy integration

- Name:web_search
- Description: Perform a web search using DuckDuckGo and generate AI-powered summaries
- Parameters:

- query(string, required): The search query to perform
- mode(string, optional): Response mode - "summary" or "detailed" (default: "summary")

Web-Scout provides MCP functionality over HTTP, accessible at the/mcpendpoint.

The MCP endpoint will be available athttp://localhost:8000/mcp

# Run the HTTP server with Docker (MCP available at /mcp) docker-compose up --build

docker run -p 8000:8000 -e GEMINI_API_KEY=your_api_key_here web-scout

To use Web-Scout as an MCP server with AI tools like Claude Desktop or Roo:

{ "mcpServers": { "web-scout": { "command": "python", "args": ["-m", "uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000"], "env": { "GEMINI_API_KEY": "your_gemini_api_key_here" } } } }

-

Configure your AI toolto use the MCP configuration:

- For Claude Desktop: Add to~/Library/Application Support/Claude/claude_desktop_config.json
- For Roo: Add to the appropriate configuration file

Can you search for the latest news about artificial intelligence?

The AI tool will use the Web-Scout MCP server (via the/mcpendpoint) to perform the search and provide summarized results.

# Install dependencies pip install -r requirements.txt # Set your API key export GEMINI_API_KEY=your_api_key_here # Run the application uvicorn main:app --reload

# Build and run docker-compose up --build # Run in background docker-compose up -d --build # Stop the application docker-compose down # View logs docker-compose logs -f web-scout

- GEMINI_API_KEY: Your Google Gemini API key (required)

- Port: 8000
- Container name: web-scout
- Health check: Automatic health monitoring

- The.envfile is ignored by Git and should never be committed
- API keys are mounted securely via Docker Compose volumes
- The application uses health checks for monitoring

- Returns proper HTTP status codes
- Includes detailed error messages
- Handles missing API keys and invalid parameters gracefully

Adds powerful web scraping and search capabilities to LLM clients like Cursor and Claude.

Scrape, crawl, and extract data from any website using the Firecrawl API.

Give Claude, Cursor, ChatGPT, Kilo, and other MCP clients access to scraping, web search, screenshots, and network tools.

Search, extract, crawl, map, and research the web — from any AI agent or terminal.

A server for web scraping, Google searches, and website URL lookups using the Olostep API.

AI tools for web scraping, crawling, browser control, and web search via the Oxylabs AI Studio API.

Web scraper exposing 5 MCP tools — crawl, markdown extraction, search-and-crawl, sitemap parsing, and LLM JSON-schema structured extraction. AGPL-3.0.

Performs web searches and extracts full page content from search results.

A Google search server using Playwright for parallel keyword searches.