Skip to main content
← Back to Frontier & Future AI
High School Lab

Prompt Engineering Lab

Engineer precise prompts to hit a target output — role, context, constraints, and format.

Prompt Engineering Lab

Prompt engineering is the skill of writing instructions that get great AI results. Choose a goal, write a prompt to achieve it, then see how your prompt scores on four real dimensions. Iterate until you reach 80+!

Choose a goal

Write your prompt

Reminder — your goal: You need an AI to explain what machine learning is. Your audience is a 10-year-old who loves video games but has never heard of AI training.
0 words0 characters

How does it actually work?

The scorer in this lab uses transparent keyword and length heuristics— no AI model is involved. Each dimension checks for specific signals: does your prompt contain numbers? Does it name an audience? Does it say “numbered list” or “avoid jargon”? Real AI systems respond similarly: they are very sensitive to these structural cues.

Prompt engineering in the real world is a loop of write → test → evaluate → revise. Professional prompt engineers keep notes on what works, build libraries of reusable patterns, and often test dozens of variations. The four dimensions here — specificity, context, constraints, and output format — map directly onto the mental checklist experts use when crafting prompts for GPT-4, Claude, Gemini, and other large language models.

Scoring is computed entirely in your browser — no network calls, no external API usage.