Prompt Engineering (CRAFT)
The quality of what you get from AI depends entirely on what you ask. The CRAFT framework gives educators a systematic approach to writing prompts that produce useful, relevant results.
Prompt engineering is the art and science of communicating effectively with AI. Just as a well-phrased question to a student elicits deeper thinking, a well-crafted prompt produces more useful AI output. For educators, this isn't about technical complexity — it's about clear communication.
Context
Set the scene. Tell the AI who you are, what subject you teach, what level your students are at, and any relevant constraints.
Role
Assign a role to the AI. 'Act as a curriculum designer' or 'You are an assessment specialist' focuses the AI's response.
Action
Specify exactly what you want the AI to do. Be precise: 'Create 5 multiple-choice questions' is better than 'Make a quiz.'
Format
Define the output format. 'Provide a table with columns for...' or 'Write in bullet points' or 'Create a rubric with 4 levels.'
Tone
Set the appropriate tone. 'Use encouraging language suitable for 8th graders' or 'Write in formal academic style.'
The CRAFT framework is flexible — you don't always need all five elements. A quick task might only need Action and Format. A complex task might use all five plus follow-up prompts. The key is intentionality: think about what you need before you type.
Structured prompting consistently outperforms unstructured requests in research. The CRAFT framework provides an easy-to-remember structure that works across AI platforms.
Teaching prompt engineering also teaches critical thinking. When students learn to craft precise prompts, they're learning to articulate their thinking clearly.
CRAFT changed my workflow. I went from spending 20 minutes tweaking AI outputs to getting what I need in one or two prompts. It's the most practical AI skill any educator can learn.
Comprehensive AI training designed for educators, by educators. From awareness to mastery.