PromptBeginnerPrompting

Prompts for learning something new

Prompts for using an assistant as a tutor rather than an answer machine, including the ones that make it stop giving you the answer.

2 min read

The second prompt ("make it ask, not tell") is the one worth trying even if you skip the rest. Reading an explanation feels like learning and mostly is not; being asked a question you cannot answer is where it actually happens.

Guard rail

Everything in this pack is about explanation, not about facts. For anything where being wrong matters (a dosage, a legal requirement, a security practice, a statistic you plan to repeat), check the underlying source. A tutor that is fluent and wrong is more convincing than one that is hesitant and right, and that is exactly the risk. See Checking Grok's answers.

What changes

Nothing here depends on a specific model. The socratic prompt occasionally drifts back into lecturing after several turns; restating "one question at a time, no explanations" puts it back.

Copy-ready prompts

4 prompts. Open one to read it, or take the whole pack.

1Explain at the right levelExplanations are either too basic or assume too much.

Explain <topic> to me.

What I already understand: <the nearest thing you do know>
What I do not: <where it goes fuzzy>
What I want to be able to do afterwards: <concrete task>

Build from what I know to what I don't. Use one example, carried all the
way through. Stop and check my understanding once, halfway, with a
question.

2Make it ask, not tellYou want to actually learn it, not read it.

I want to learn <topic>. Do not explain it up front.

Ask me one question at a time, starting from what I likely already know.
After each answer, tell me briefly whether I am right and why, then ask
the next question. When I get something wrong, do not just correct it, ask a smaller question that shows me why it is wrong.

Keep going until I can explain <the specific thing> without help.

3Find the gap in my understandingSomething works and you cannot say why, or does not and you cannot say why not.

Here is my current understanding of <topic>:
"""
<write it out in your own words, badly if necessary>
"""

Point out specifically: what is wrong, what is right but for the wrong
reason, and what is missing entirely. Be direct about the wrong parts, do not soften them. Then give me the single correction that would fix the
most of it.

4Worked example, then a variationYou need to see the shape of a solution before trying one.

Show me a worked example of <task>, narrating the decision at each step, not just what you did but why that and not the alternative.

Then give me a second problem that is the same shape but different
enough that copying the first answer will not work. Do not solve it.

Afterwards I will send you my attempt; review it against the reasoning in
the worked example, not against your own preferred solution.

Last checked against xAI’s own pages on 2026-08-20. Grok changes quickly; anything version-specific should be confirmed upstream before you rely on it.

Board pack narrative for director review

Draft a board-ready performance narrative from management accounts and operating metrics, for finance directors and FP&A leads.

Prompting7 min read

WorkflowIntermediate

Earnings summary for investor communications

Produce a sourced earnings briefing from results, guidance and analyst questions for investor relations and senior management.

Prompting6 min read

WorkflowIntermediate

Model documentation for review handover

Produce a review-ready model documentation pack from a workbook, source data and change notes for finance teams.