What problem does it solve? Users often accept confident, well-structured AI answers at face value, even when those answers contain estimates, advice, or claims they will act on. This Skill adds a brief moment of reflection after substantive answers, helping users catch bad assumptions or missing context before it matters. ## Core Features & Use Cases - Targeted follow-up prompts: Generates 2-3 short questions tied to specific facts, reasoning steps, or assumptions in the answer just given. - Clear activation boundaries: Applies only to substantive, actable content such as advice, estimates, drafted plans, and data interpretation, while skipping trivial lookups, educational explanations, code, and creative writing. - Once-per-conversation limit: Prevents the nudge from becoming repetitive by offering it at most one time per conversation. - Use Case: After receiving a marketing budget plan with CPL estimates and a 70/30 channel split, the user gets prompts like "How do these CPL estimates compare to benchmarks in my specific vertical?" to verify the plan before acting on it. ## Quick Start Ask the assistant to review its substantive answer and append a few specific follow-up questions that help you verify key facts and assumptions before acting on it.
