The Video Analyzer is a powerful tool designed to dissect video content, transforming it into a rich, structured storyboard breakdown. It intelligently extracts keyframes using scene detection, transcribes audio for timestamped narration, and leverages Claude Vision for in-depth visual analysis. Beyond basic breakdown, it computes an 8-field stylistic fingerprint, providing a deterministic classification of various video attributes, making it invaluable for content analysis, creative review, and detailed documentation. It can be used standalone or integrated as an MCP server for enhanced functionality within Claude Desktop or Claude Code.
Key Features
01Combined shot-by-shot storyboard output (.docx or .md)
02Scene-detection-based keyframe extraction
03Per-frame visual descriptions using Claude Vision
04Timestamped audio transcription via OpenAI Whisper
058-field Stylistic Fingerprint v3 classification
060 GitHub stars
