How rea-video Reverse-Engineers Any Video Shot by Shot
GitHub Open Source
Reverse-engineer any MP4 into a shot-by-shot remake plan separating live footage, AI generation, and code.
Repository snapshot
- Repository: JuneYaooo/rea-video
- Language: Python
- License: Apache-2.0
- Stars: 5 499
- Forks: 789
Daily open-source picks: GitHub Open Source on Telegram
What it is
rea-video is a command-line utility and agent skill designed to reverse-engineer any video file shot by shot. It determines whether each segment was filmed with a camera, generated by an AI model, or rendered through code and templates. Built around Python and FFmpeg, it acts as an analytical layer that plugs directly into developer environments like Claude Code and Codex.
Why it matters
Deconstructing modern mixed-media videos usually requires tedious manual inspection across multiple editing and diagnostic tools. By automating frame extraction, audio timing, and synthetic artifact detection, this project cuts down the reconnaissance phase of recreating a video from hours to seconds. It provides clear technical breakdowns rather than vague visual guesses.
Key features
- Classifies every shot as filmed, AI-generated, or coded with duration weighting.
- Generates prompt recipes and camera notes for recreating each segment.
- Exports time-coded captions and styled text as SRT or ASS files.
- Parses editing rhythm, cuts, audio tracks, and background music BPM.
- Integrates directly into AI coding agents through an installable skill.
Quick start
To install the tool and register its agent skill, run the following commands in your terminal:
brew install ffmpeg
git clone https://github.com/JuneYaooo/rea-video && cd rea-video
bash scripts/install_skill.shWho should try it
This tool fits developers, digital artists, and technical creators who frequently study reference videos to build similar motion graphics, UI demos, or mixed-media content.
Verdict
rea-video offers a pragmatic approach to video analysis that saves substantial manual effort when breaking down complex edits. It is a specialized utility that earns its place if you routinely reverse-engineer reference material.
Continue exploring
Get the next pick in Telegram: Follow GitHub Open Source
Curated from the repository README and metadata. No hands-on test is claimed; evaluate it in a sandbox before production use.