Your question is Evaluate Medical Advice Response Quality. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are reviewing how an assistant should be evaluated when users ask for medical advice. The team wants a clear framework for judging whether responses are high quality, appropriately toned, and factually accurate, especially in higher-risk health conversations.
What metrics would you use to measure the quality, tone, and accuracy of Claude's responses in a medical advice scenario?