Prompt-Craft Cards: A Toolkit for Developing Design Judgment through Reverse Prompt Engineering

Abstract

The integration of generative AI into design education poses a critical challenge beyond technical skill: cultivating design judgment. Students often struggle to externalize their tacit knowledge into the explicit language of prompts, a process essential for developing the analytical capacity to make well-substantiated design choices. This paper introduces Prompt-Craft Cards (PCC), a tangible toolkit and pedagogical framework designed to address this gap. Using a Research-through-Design methodology, PCC frames Reverse Prompt Engineering (RPE) as a simulator for practicing design decisions at specific expertise levels. The toolkit features a differentiated system of three card decks, each scaffolding a different stage of expertise and knowledge transfer based on established learning theories. The Foundational Deck targets compositional judgments, the Adaptation Deck focuses on navigational choices, and the Reflection Deck encourages critical, metacognitive inquiry. The toolkit's efficacy will be evaluated through a multi-stage study employing think-aloud protocols within an Action Research design. This structured process transforms prompting into a form of deliberate practice. By engaging with the ambiguity between their intent and the AI's output, students are compelled to articulate, reflect, and refine their design decisions. Ultimately, I argue that this structured, reflective practice moves beyond teaching prompt engineering as a technical skill, transforming it into a powerful pedagogical method for cultivating the critical visual literacy and metacognitive thinking essential for professions and practices that relies on interpreting and creating visual media in the age of AI.