Development · GitHub repo
animation-reverse-engineering
Claude Code skill to turn animation videos into production Framer Motion code.
This Claude Code skill extracts exact timing, easing, and stagger parameters from video references like X clips, Dribbble shots, or screen recordings. It uses ffmpeg and yt-dlp to dissect frames instead of guessing transitions, then writes the corresponding Framer Motion implementation. It supports both timeline-driven and interaction-driven animations.
What it's for
- Replicate UI animation references from X or Dribbble into Framer Motion code
- Extract frame-by-frame timing and easing parameters from screen recordings
- Analyze interaction-driven animations to map gesture triggers and progress curves
Features
- Automated video acquisition via yt-dlp and ffprobe
- Frame dissection generating 2fps contact sheets and crop-band stacks at native fps
- Classification pipeline separating timeline-driven and interaction-driven animations
- Framer Motion easing and stagger extraction derived directly from frame deltas
- Interactive gesture grammar and progress mapping analysis
- Command-line dissect script for manual extraction workflows
Pricing
Free and open source (MIT license)
Details
- Type
- GitHub repo
- Category
- Development
- Topic
- AI Coding Tools
- Open source
- Yes
- Platforms
- Command line, macOS, Linux, Windows
- For
- Front-end developers and creative technologists building web animations with Framer Motion.
- Tags
- claude code, framer motion, animation, reverse engineering, ffmpeg, web development
Why it was saved
- @0xsrinivasa: Makes replicating cool animations seen on X practical: the skill measures frame-level timing instead of guessing, and classifies animations as timeline-driven vs interaction-driven to generate accurate easing, stagger order, progress mappings and gesture triggers.
Related
Design inspiration MCP server and archive for AI coding agents.
Curated skills and performance rules for AI coding agents by Vercel.
Curated roster of 20 AI app builders and coding assistants.
Design skill preventing LLMs from generating generic, AI-looking web interfaces
Walkthrough for setting up Model Context Protocol servers inside Cursor
Unified Mac HUD to triage and control multiple coding agents
Updated