An automated pipeline that leverages LLM's meta-learning capability to iteratively design and refine red-teaming systems without human intervention. - View it on GitHub