ButterCut addresses the time-intensive nature of video editing by integrating Claude Code, an AI assistant, to automate the creation of roughcuts and edit timelines. It targets video professionals and enthusiasts seeking to leverage AI for accelerated post-production workflows, offering a significant reduction in manual editing effort by generating editor-ready sequences for Final Cut, Premiere, and Resolve.

How It Works

The project employs a novel approach where Claude Code, guided by ButterCut Skills and a Ruby library, analyzes video footage. It leverages tools like FFmpeg for media processing and WhisperX for audio transcription. The system generates comprehensive visual and audio transcripts, enabling Claude to make editorial decisions based on user-defined preferences for narrative structure, duration, and pacing. This AI-driven analysis culminates in the export of FCPXML timelines directly compatible with professional non-linear editing software.

Quick Start & Requirements

• Installation: Clone the repository (git clone https://github.com/barefootford/buttercut.git && cd buttercut) and run claude or claude --dangerously-skip-permissions within the project directory, then instruct Claude to "Install ButterCut".
• Prerequisites: Ruby, Python, FFmpeg, and WhisperX. Claude automatically checks for and installs missing dependencies. Detailed manual installation instructions are available in docs/installation.md.
• Resources: Processing involves significant analysis and transcription, with no specific time or resource footprint estimates provided.
• Documentation: Installation guide (docs/installation.md), basic XML generation (docs/basic-xml-generation.md), and demo videos are mentioned.

Highlighted Details

• Interactive video editing powered by Claude Code.
• Automated library creation, audio transcription (WhisperX), and visual frame analysis.
• Direct export of timelines for Final Cut Pro X, Adobe Premiere Pro, and DaVinci Resolve.
• User-configurable editorial parameters including narrative structure, duration, and pacing.

Maintenance & Community

The project was inspired by the work of Chris Hocking and Andrew Arrow. Contributing guidelines are provided, emphasizing self-authored pull request descriptions and small, focused contributions. No specific community channels (e.g., Discord, Slack) or roadmap links are detailed in the README.

Licensing & Compatibility

The project is released under the MIT license, which is highly permissive and generally compatible with commercial use and integration into closed-source projects.

Limitations & Caveats

The core functionality is dependent on the Claude AI platform. The use of a --dangerously-skip-permissions flag suggests potential system access considerations. The README does not specify an alpha or beta status, but the reliance on external AI services and interactive setup implies an experimental nature.