WebSearch Tool

Example Use Cases

• 1

  Perform targeted web searches and return structured search results from live web sources.

• 2

  Extract specific information from multiple webpages via natural-language prompts and include source references.

• 3

  Crawl or scrape websites (with depth and page limits) to collect content, screenshots, or markdown/HTML output for downstream analysis.

Description

WebSearch is a Python-based MCP server that leverages the Firecrawl API to provide advanced web search, site crawling, scraping, and targeted content extraction. It can extract structured information from lists of URLs using natural-language prompts, optionally augmenting results with OpenAI or Tavily. The project includes configurable crawl depth/limits, scraping output options (markdown/HTML), and integration instructions for Claude for Desktop. It is aimed at agents that need reliable web content retrieval and analysis while relying on external API keys for operation.