This repository provides an AI-powered video clipping pipeline that processes long-form YouTube videos into viral-ready short clips. It leverages GPT-class LLMs to identify and rank the most engaging moments based on a customizable virality framework, and uses OpenAI’s Whisper model for audio transcription. The system then auto-crops the selected highlights into 9:16 vertical format optimized for TikTok, Reels, and YouTube Shorts. The project is self-hostable, requires no per-clip credits or watermarks, and can be used as a Python library or CLI tool.