video-expert-analyzer

Automated video analysis scores scenes and selects the best shots by applying Walter Murch's Six Rules, requiring multimodal vision models and a configurable pipeline to generate scene scores and analysis reports.

Scene detection and scoring supports videos from Bilibili, YouTube, Douyin, and Xiaohongshu by processing URLs through agent or API modes to generate comprehensive analysis reports.

Automated video scene scoring requires multimodal vision models such as Gemini 3.0, Kimi 2.5, or Claude with vision to evaluate scenes and identify standout moments.

Batch processing large video projects uses a configurable pipeline to extract subtitles, detect scenes, and generate comprehensive analysis reports with a bilingual glossary.

Curating top clips from long videos uses automated scoring workflows that apply Walter Murch's Six Rules to evaluate scenes and surface standout moments, saving editors hours of manual review.

Batch video analysis works with both agent and API modes for professional teams, processing large projects to produce scene scores, best shots, and final analysis reports.