メディア
Video Transcriber
Transcribes videos from YouTube, Facebook, Instagram, TikTok, and other platforms to clean
markdown text. Uses a 3-tier strategy: YouTube caption API (fastest), yt-dlp subtitle
extraction (any platform with captions), and Whisper fallback (universal).
Use this skill whenever the user wants to: transcribe a video, get text from a YouTube/Facebook/
Instagram/TikTok video, extract captions or subtitles, convert a video to text, get a
transcript, or turn a video into readable text. Also triggers on phrases like "what did they
say in this video", "summarize this video" (transcribe first, then summarize), or any request
involving video URLs where the user needs the spoken content as text.