Media Editor

An MCP (Model Context Protocol) server that provides tools for editing media files. Use it with Claude or other MCP-compatible AI assistants to automate video and audio processing tasks.

Features

• Video Trimming - Cut video files to specific time ranges
• Audio Transcription - Convert speech to text using NVIDIA NeMo
• Themed Thumbnails - Generate and prepend title cards to videos with customizable themes

Prerequisites

• Python 3.12+
• uv package manager
• FFmpeg (must be installed and available in PATH)
• NVIDIA GPU with CUDA support (for transcription)

Getting Started

Installation

This repository uses Git LFS to store test video files. Install Git LFS before cloning:

# Install Git LFS (macOS)
brew install git-lfs

# Initialize Git LFS
git lfs install

Then clone and install:

# Clone the repository (LFS files are fetched automatically)
git clone https://github.com/wmeints/media-editor.git
cd media-editor

# Install dependencies
uv sync

Verifying Your Setup

Run the doctor command to check that all required dependencies are installed and available:

uv run media-editor doctor

If successful, you'll see the path to FFmpeg. If FFmpeg is missing, install it and ensure it's in your PATH.

Running the MCP Server

uv run media-editor mcp

Connecting to Claude Desktop

Add the following to your Claude Desktop configuration (claude_desktop_config.json):

{
  "mcpServers": {
    "media-editor": {
      "command": "uv",
      "args": ["run", "media-editor", "mcp"],
      "cwd": "/path/to/media-editor"
    }
  }
}

Development

# Run tests
uv run pytest

# Type checking
uv run pyright

# Linting
uv run ruff check src/
uv run ruff format src/

License

MIT