Reliable Decision Support with LLMs: A Framework for Evaluating Consistency in Binary Text Classification Applications