Hands-on · Workshop 02

Prompting for Real Work

A vague request gets a vague answer, and most of the time you blame the model when the real gap was in what you handed it. This workshop runs one real request through a repeatable loop: frame what you actually want, specify the context and constraints the model cannot guess, show one example of the output shape, decide how you will judge the result before you read it, then own the result and edit it to done. You leave with a reusable prompt structure, a review rubric you write first, and a before-and-after of your own request.

Duration

45 to 60 minutes

Level

All levels

Tools

A few real requests you have sent recently, a doc or whiteboard, and an AI assistant to test your rewrites

Outcome

A repeatable way to write prompts that get usable results, and to judge whether the result is actually usable