Enables LLMs to interact with YouTube videos by fetching transcripts, summarizing content, and answering questions based on video context.