Explore. Preview. Adapt. Five ready-to-use activities. Built for your classroom.
👆 Drag to explore • Click a card to preview
Explore. Preview. Adapt. Five ready-to-use activities. Built for your classroom.
👆 Drag to explore • Click a card to preview
None of these activities required me to be a programmer. Each one was designed, directed, and curated by a faculty member and built in conversation with AI. Two ideas make that possible — and both are worth teaching in their own right.
Context engineering is the deliberate practice of shaping what you give an AI — the role, audience, evidence, examples, format, and constraints — so its output actually fits your purpose. The model is only as good as the context a thoughtful human supplies.
In academic settings it reframes "prompting" from a magic phrase into a teachable rhetorical skill. The moves students use to engineer good context — naming an audience, marshaling evidence, setting boundaries — are the same moves that underlie strong academic writing and critical thinking.
This gallery's name, Context Is All You Need, is a nod to exactly that: better context, not bigger models, is what makes AI genuinely useful for learning.
Vibecoding — a term popularized by Andrej Karpathy — means building software by describing what you want in plain language and letting AI generate and refine the code, staying in a creative flow instead of writing every line by hand.
For faculty, it collapses the distance between a teaching idea and a working tool. You can prototype an interactive activity in an afternoon, test it with students, and revise it — keeping pedagogy, disciplinary expertise, and human judgment firmly in the loop while the AI handles the scaffolding.
Every experiment here was vibecoded. The takeaway for students isn't "let AI do it" — it's that clear thinking, well expressed, is what builds things.