MCPDocSearch is a tool for crawling websites, generating Markdown documentation, and making it searchable via an MCP server. It consists of two main components: a crawler CLI that extracts content from documentation sites and converts it to Markdown, and an MCP server that loads these documents, chunks them into searchable sections, and generates vector embeddings using sentence-transformers. The server exposes tools for listing documents, retrieving document structures, and performing semantic searches over the content, making it particularly useful for AI agents that need to access up-to-date documentation beyond their training cutoff.
暂无评论. 成为第一个评论的人！
登录以参与讨论
Lists available crawled documents.
Retrieves the heading structure for a document.
Performs semantic search over document chunks using vector similarity.
