Web Scout

Provides web search capabilities with AI-powered summaries via an HTTP MCP endpoint

python

0

GitHub Stars

python

Language

6 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

Web-Scout is an MCP server that lets AI assistants perform web searches and receive AI-generated summaries or detailed analyses. It exposes an HTTP MCP endpoint for integration and can also run as a local process, enabling seamless, context-rich web search capabilities within your tooling ecosystem.

How to use

You connect your MCP-enabled AI assistant to the Web-Scout MCP endpoint to perform web searches and obtain AI-powered summaries. Use the web_search tool to submit a query and choose a response mode. When using the HTTP MCP endpoint, the server responds with a structured result that includes the query, the chosen mode, a concise summary or detailed analysis, and the number of sources consulted.

How to install

Prerequisites: Ensure you have Docker and Docker Compose installed. You also need a Gemini API key for AI-powered summarization.

Step by step install and run using Docker Compose:

Prepare environment variables in a file named .env placed in your project directory and include your Gemini API key.
Build and run the container stack with Docker Compose.
Access the MCP endpoint at http://localhost:8000/mcp and the HTTP server at http://localhost:8000.

If you prefer running the server directly without Docker, install Python dependencies and start the app with Uvicorn.

Additional sections

Configuration and environment considerations: The Gemini API key must be provided as GEMINI_API_KEY in your environment. When running via Docker, pass this value through the environment or a .env file that Docker Compose references. The server exposes the MCP endpoint at /mcp for integration with AI tools.

Security notes: Do not commit your Gemini API key or any credentials. Use environment variables and secure secret management where possible. The HTTP MCP interface uses JSON-formatted responses suitable for integration with AI tooling.

Troubleshooting tips: If health checks fail or the MCP endpoint is unreachable, verify that the GEMINI_API_KEY is set correctly, check container logs for startup errors, and ensure the server is listening on port 8000 as configured.

Available tools

web_search

Perform a web search and generate AI-powered summaries or detailed analyses using DuckDuckGo for search and Gemini AI for summarization