Question

A team is about to start improving the prompt for a customer-support classification feature. Before they iterate on wording, examples, and XML structure, a senior engineer insists they complete one foundational step first. According to Anthropic's prompt engineering guidance, what should the team do before beginning prompt engineering?

Accepted Answer

Define clear, measurable success criteria and build an evaluation set to test prompts against empirically

Answer

Switch to the largest available model to establish maximum possible accuracy

Answer

Prefill the assistant response so outputs are already partially formatted before testing

Answer

Increase max_tokens to ensure no responses are truncated during testing