The Context-First Learner
Start every new AI account with this. It sets the foundation for everything that follows.
I'm a [your role, e.g. college sophomore] studying [subject] at a [level] level. I want to deeply understand [specific topic]. Here are my current notes and materials: [paste notes]. Take this information and explain it back to me in a way that makes sense. Ask me clarifying questions if anything is unclear about my level or goals.
The AI Tutor Quiz
Turn ChatGPT into a free, personalized tutor that adapts to your weak spots in real time.
Here are my study notes on [topic]: [paste notes]. Quiz me one question at a time. Wait for my answer before moving on. After each answer, tell me if I'm right or wrong and explain why. Track which areas I struggle with and circle back to them. Adjust difficulty based on my performance. Let's start.
The Feedback Trainer
Feed the AI your best work so it learns your standards. This is how you calibrate it to YOU.
Here is a piece of work I'm proud of: [paste your work]. This represents my standard of quality. Here's what I liked about it: [what worked]. Here's what I'd improve: [what you'd change]. Remember this as my benchmark. When I ask for help in the future, aim for this level of quality and match this writing voice.
The Rubric Grader
Get feedback that actually matches what your professor wants — not generic "good writing" standards.
Here is my [essay/assignment]: [paste work]. Here is the rubric my professor uses: [paste rubric]. Notes about my professor's preferences: [what they emphasize, penalize, reward]. Grade this work using ONLY the rubric criteria. Give me a letter grade first with no explanation. Then I'll ask for feedback.
The Interview Destroyer
Know every hard question before you walk into the room. Two AIs. One hour of prep. Unstoppable.
Here is the job description: [paste job description]. Here is my resume: [paste resume]. Based on the gaps between this role's requirements and my experience, give me the 10 hardest questions the interviewer is likely to ask. Be ruthless — I'd rather be over-prepared than caught off guard. For each question, explain why they'd ask it and what they're really testing.
The Meeting Stress-Tester
Ask AI to destroy your pitch before your client does.
I have a meeting with [client/stakeholder description]. Here are my prep notes: [paste notes]. Here is the proposal I'm presenting: [paste or summarize]. Act as a skeptical version of this client. Give me the 5 hardest questions they could ask. Then identify any weak arguments, unsupported claims, or assumptions I'm making that I might not realize.
The Weekly Planner
Dump your chaos into AI every Sunday night and get back a clean, prioritized weekly plan.
Here's everything on my plate this week: [list all tasks, deadlines, meetings, personal commitments]. My working hours are [hours]. I do my best deep work in the [morning/afternoon]. Build me a realistic day-by-day schedule with time blocks. Include buffer time between tasks. Rank priorities. Flag anything that seems unrealistic given the time available.
The Doctor Translator
After any appointment, paste what you remember and get it explained in plain language.
I just had a doctor's appointment. Here's what I remember being told: [everything you recall — terms, medications, instructions]. Please explain each medical term in plain language. Summarize the key takeaways I should remember. List any follow-up questions I should ask at my next appointment. Note: this is for my understanding only, not a replacement for medical advice.
The Code Explainer
Never use code you don't understand. Force the AI to teach you, not just give you the answer.
Here is code I need to understand: [paste code]. Walk me through this line by line. For each section, explain: what it does, WHY this approach was chosen over alternatives, and what would break if I removed or changed it. Use analogies where possible. I'm at a [beginner/intermediate/advanced] level.
The Bug Describer
Describe the problem in English before asking for the fix. Half of debugging is articulating the issue.
I'm working on [project description] using [language/framework]. I have a bug. Here's what SHOULD happen: [expected behavior]. Here's what ACTUALLY happens: [actual behavior]. Here's the relevant code: [paste code]. Before giving me a fix, explain what you think is causing the issue and why. Then show me the solution and explain the reasoning.