iopho-analyzing-videos — Video → .storyboard.md Reverse Engineering

Analyze a video file and generate a detailed .storyboard.md file — a human+AI readable "source code" representation of the video, including scene-by-scene breakdowns, visual descriptions, audio transcripts, and extracted keyframes.

Prerequisites

!python3 -c "import google.generativeai; print('google-generativeai: ✓')" 2>/dev/null || echo "google-generativeai: ✗ — install: pip install google-generativeai"

!which ffmpeg 2>/dev/null && echo "ffmpeg: ✓ (needed for frame extraction)" || echo "ffmpeg: ✗ — install: brew install ffmpeg"

!echo "GEMINI_API_KEY: $([ -n \"$GEMINI_API_KEY\" ] && echo '✓ set' || echo '✗ — set via: export GEMINI_API_KEY=your-key')"

Required: google-generativeai + ffmpeg + GEMINI_API_KEY environment variable.

Get a Gemini API key at https://aistudio.google.com/apikey (free tier available).

What It Produces

Given a video file, this skill generates: