On the Brittle Foundations of ReAct Prompting for Agentic Large Language Models

Open in new window