A Model Context Protocol (MCP) server that extracts content from multiple video platforms (Douyin, Bilibili, Xiaohongshu, Zhihu) and generates intelligent knowledge graphs with OCR support.