GOOD PROMPTS ARE DESIGNED, TESTED AND REVISED
Say what you need.
Make it testable.
Turn vague requests into clear, testable instructions. Study prompt makeovers, repair weak briefs, and build a library of prompts you can explain.
CHANGE AN INPUT / INSPECT THE RESULT
Your prompt workshop.
Build, challenge and revise your instructions. Workshop drafts save with your progress. No API key or live model is connected; the checks make their limits explicit.
TURN IDEAS INTO TESTABLE WORK
Build a prompt worth reusing.
Download a project brief, record your decisions and collect real evidence. Package your brief, versions, test cases and review evidence.
RETURN / RECALL / REVISE
Keep what you understand.
Recall practice
Your learning record
Completion requires a correct scenario answer, an artifact of at least 80 characters and a reflection of at least 40 characters. Editing an answer or written work reopens the lesson.
Review timing follows your own rating, not a measured estimate of mastery.
Explore primary sources
NIST AI Risk Management Framework ↗NIST Generative AI Profile ↗Anthropic: Building effective agents ↗Anthropic: Evaluating agents ↗Anthropic: Effective context engineering ↗DiscoveryVIP AI Learning Studio ↗DiscoveryVIP JavaScript Studio ↗This independent course teaches practical prompt design and evaluation. Completion records practice, not independently verified mastery or provider certification. Product capabilities, prices and policies change; verify them with the provider you choose.
Use fictional data. Your drafts and progress live in this browser, and exports contain your written work. The page includes the website’s Google analytics tag; it does not attach your practice text to custom analytics events.