A modular PySide6 desktop app to automate image dataset labeling with vision models (GPT-4o, Gemini). Features a unique visual node-based parser and placeholder system to dynamically generate highly customized prompts from filenames and folder structures, perfect for training diffusion models.