Web MCP Server logo

Web MCP Server

by zgagnon

Web MCP Server provides web search and content retrieval capabilities for AI assistants. It offers tools for web searching and extracting content from URLs.

View on GitHub

Last updated: N/A

What is Web MCP Server?

A Model Context Protocol (MCP) server that enables AI assistants to perform web searches and retrieve content from web pages. It provides tools to search the web using DuckDuckGo and extract readable content from URLs.

How to use Web MCP Server?

The server can be run locally using npx https://zgagnon.com/webmcp or nix run github:zgagnon.com/webmcp. It exposes an API with two main endpoints: search_duckduckgo for web searches and read_url for fetching and processing content from URLs. Each endpoint takes specific input parameters and returns structured output.

Key features of Web MCP Server

  • search_duckduckgo tool

  • read_url tool

  • HTML markup removal

  • Mozilla's Readability integration

  • Schema validation with Zod

Use cases of Web MCP Server

  • Enhancing AI assistant knowledge

  • Providing real-time information to AI models

  • Automated content extraction for AI training

  • Web research for AI agents

FAQ from Web MCP Server

What is the purpose of the search_duckduckgo tool?

It performs web searches using DuckDuckGo and returns a list of URLs matching the search query.

What does the read_url tool do?

It fetches the content of a specified URL, removes HTML markup, and extracts the main content using Mozilla's Readability.

What are the input parameters for search_duckduckgo?

It requires a 'query' (string) and optionally accepts a 'limit' (number) for the maximum number of results.

What are the input parameters for read_url?

It requires a 'url' (string) and optionally accepts 'useReadabilityMode' (boolean) and 'maxLength' (number).

What happens if an error occurs?

The server will return an error if invalid arguments are provided, URL fetching fails, or search execution fails.