Interactive Concept Comprehension Check

When you ask an AI to "explain redox reactions," you get a beautifully organized summary, and while reading it, you feel like you understand everything. However, in front of an actual exam sheet, those sentences vanish. Reading comprehension and active retrieval are entirely different cognitive skills. By flipping the dynamic so the AI tests your understanding, the gap between what you actually know and what you merely recognize appears on the very first question.

Setting the Role in the first paragraph as "your job is to ask questions, not explain" is critical because language models default to lecturing. They are inherently eager to assist, so if given leeway, they will kindly attach the answer right after the question. Hardcoding "do not explain" inside the role is what keeps the back-and-forth dialogue alive.

The fourth paragraph establishes the rules for Follow-up Questioning. Breaking down a roadblock into smaller guiding questions rather than handing out the answer is the most vital instruction in this prompt. This mimics what a great human tutor does, helping the learner pinpoint the exact step where their logic failed. Allowing an explanation after three consecutive failed attempts prevents endless frustration and abandonment—a common failure mode of AI tutor prompts lacking an exit ramp.

The Constraints in the fifth paragraph ("one at a time" and "ask only the first question") ensure an interactive loop. Without these constraints, models dump a list of ten questions at once, killing the conversation. Dynamic difficulty adjustment only works when exchanging one question at a time. Finally, the "Stop" trigger wraps up the session into an actionable study checklist.

The AI outputs ten questions along with their answers and explanations all at once. Because the answers are visible right below the questions, your eyes naturally wander, leaving you with a false sense of mastery. Since the difficulty is static, you cannot identify where your understanding breaks down, and without anyone probing why you made a mistake, misconceptions remain untouched.