The Prompting Cookbook: Shifting from Prompting to Context Engineering

Your students are taking your class, not a generic, average internet class. Context engineering is the practice of putting your own values, course materials, and academic standards at the center of what an AI responds with — instead of letting the open web fill in the answer. One-line prompting tricks are giving way to this more deliberate approach: anchoring AI output in the specific sources you choose. Using NotebookLM and comparable tools, you'll practice grounding AI directly in your own syllabus and readings, and see why a well-fed model behaves very differently from an unanchored chatbot. The session closes by flipping the lens — context engineering is a student skill, too — and you'll design an assignment component that asks students to curate and document the sources they feed to AI. The 30-minute work session is for grounding a tool in your own course materials.