Video Intelligence

Generate a three-scene video, independently inspect container metadata, scene boundaries, and audio, merge the measurements into a timeline, and publish a validated report. The supplied Docker image uses FFmpeg for the actual video analysis.