Canonical: https://alcub3.com/learning/foundations/prompt-craft-101
Updated: 2026-10-24

[← Back to Foundations](https://alcub3.com/learning/fundamentals.html)
Path 01 • Module 03
#
Prompt craft that
actually travels
.
Prompting is not about secret magic phrasing. It is about reducing ambiguity and giving the system the right frame for the work.
Time
Self-paced reading + practice
Level
Beginner
Outcome
Repeatable prompt patterns
##
The four levers that matter most
Role framing:
tell the model what lens to adopt.
Specificity:
define deliverable shape, audience, and use case.
Examples:
show what “good” looks like when you can.
Iteration:
repair the answer instead of abandoning the workflow.
Request design is one source of poor output. Model limits and missing or unreliable input can also cause failures.
##
Patterns you can reuse immediately
###
For drafting
“Write a concise first draft for
[audience]
about
[topic]
. Keep it under
[length]
. Use a tone that feels
[style]
. Include three concrete points and end with a clear next step.”
###
For analysis
“Review this material and return the answer in three sections:
what matters, what is uncertain, what we should do next
.”
###
For decision support
“Compare these options using
criteria, tradeoffs, and recommendation
. Do not assume facts not stated here.”
###
Try it now
Take one recurring task you do badly under time pressure. Create a reusable prompt for it. Save the structure somewhere you can reuse, then run it again next week and improve only the weakest part.
Prompt rule
###
Specific beats clever.
Start with clear constraints and test whether they improve your result.
###
Build prompt assets.
Your best prompts should become reusable tools, not one-off experiments.
[Previous Your first AI conversation](https://alcub3.com/learning/foundations/first-ai-conversation.html)
[Next AI for your work](https://alcub3.com/learning/foundations/ai-for-your-work.html)
##
Worked example: Test the reusable prompt
For a weekly status summary, define three sections: completed work, blocked work and decisions needed. Include one example of a useful bullet. Run the same prompt on two different sets of non-sensitive notes.
###
Check your exercise
A useful template works on both inputs, keeps the requested structure and invents no missing facts. Record one failure and revise the instruction that could address it; retest rather than assuming the edit worked.
Illustrative example. Use fictional or non-sensitive inputs for practice.
Further reading:
[NIST AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework)
for risk identification and evaluation.
Published by ALCUB3 · Updated October 24, 2026.