A useful prompt library should strengthen learning, not automate every assignment. Keep a small set of patterns that ask for hints, questions, investigation steps, explanations, tests, reasoning review, quizzes, and manual verification.
Use these only when course policy and data rules permit the interaction. Replace bracketed fields without adding private or restricted material.
Ask for a hint
I am working on [problem type]. My current approach is [summary], and I am stuck
at [specific point]. Give one hint or question that helps me choose the next
step. Do not provide the complete solution.
Try the hint before asking for another. Record what changed in your reasoning.
Investigate an error
Help me investigate this error without rewriting the assignment. Here is a
minimal reproducible example and the complete error message: [safe material].
List three plausible hypotheses, the evidence each predicts, and one check I can
run manually. Preserve my approach.
Do not paste credentials, private paths, full repositories, or restricted course code. Verify hypotheses in the actual environment.
Request a concept explanation
Explain [concept] for a student who understands [prerequisite] but not
[confusing point]. Start with a precise definition, then one small example and
one boundary case. End with two questions that test understanding.
Compare definitions with the course text or another primary source.
Generate tests without surrendering design
Given this behavior specification, suggest test categories: expected, boundary,
and invalid inputs. Do not write implementation code. Explain what failure each
test could reveal: [specification].
If test design is assessed, this use may be prohibited. Check policy first.
Review reasoning
Review my reasoning for unsupported assumptions or missing cases. Do not replace
it with a polished answer. Quote the exact step that needs attention, explain the
problem, and ask me to revise it: [reasoning].
An AI critique can also be wrong. Confirm objections through calculation, evidence, tests, or instructor guidance.
Generate a quiz
Create five short retrieval questions from these verified learning objectives.
Ask one at a time. Wait for my answer, give specific feedback, and cite which
objective it tests. Do not introduce facts outside the supplied material.
Use authorized notes rather than uploading restricted lecture material.
Request manual verification
Do not tell me merely that the result looks correct. Give a manual verification
procedure with observable evidence, likely failure modes, and a clear stopping
condition. Separate checks you can infer from checks I must run locally.
This pattern counters plausible but unverified conclusions.
Store patterns, not answers
Keep prompts in Markdown with:
- intended learning purpose;
- allowed context;
- prohibited inputs;
- expected response shape;
- manual verification; and
- one example using fictional data.
Do not store course answers, private transcripts, or secrets in the library. Review patterns when course rules or tools change.
Common mistakes
- Asking for “a hint” after supplying no attempt.
- Treating prompt wording as a guarantee.
- Uploading more context than policy permits.
- Generating tests that reveal the assessed solution.
- Accepting citations without opening primary sources.
- Saving outputs instead of the reusable interaction pattern.
Do this now
Choose one allowed pattern. Write down what you know, what remains unknown, and what evidence would verify the response. Use the pattern on a safe practice problem, then perform the verification.
Log what you learned
Record only:
- Result: What did the action produce?
- Evidence: What observation, test, or source supports that result?
- Next action or unresolved question: What should happen next?
Next, turn a useful conversation into a durable learning record rather than leaving it in chat.