What problem does it solve? Converting recorded meetings and video files into searchable, speaker-labeled text transcripts normally requires manual work or paid cloud services. This Skill automates the full pipeline locally: extracting audio from video files and transcribing it with speaker diarization into structured markdown. ## Core Features & Use Cases - Audio Extraction: Pull audio tracks from .mov, .mp4, and .mkv files into MP3 using FFmpeg with draft, standard, and archival quality presets. - AI Transcription with Diarization: Transcribe audio using whisperX large-v3 with pyannote speaker diarization, producing markdown with speaker labels and timestamps. - Multiple Output Formats: Export transcripts as markdown, SRT, VTT, plain text, or JSON. - Dependency Doctor: Verify FFmpeg, whisperX, pyannote, HuggingFace token, and system specs before running. - Use Case: Record a one-hour meeting in OBS, then run the QTRANSCRIPT workflow to get a markdown transcript with Speaker 1/Speaker 2 labels and timestamps in about three minutes. ## Quick Start Transcribe the video file at /path/to/meeting.mov into a markdown transcript with speaker labels using the video transcript skill.
