Prompt Engineering Fundamentals
Learn the core principles of crafting prompts that get useful, accurate responses
What is Prompt Engineering?
Prompt engineering is the skill of crafting inputs to AI systems that produce the outputs you need. It's not about tricking AI or finding magic words—it's about clear communication and structured thinking.
Think of it like giving instructions to a highly capable assistant who has never met you before. The clearer and more specific your instructions, the better the results.
The Five Core Principles
1. Be Specific
Vague prompts produce vague results. Instead of “Help me with my lesson,” try “Help me create discussion questions for a Year 10 history lesson on the Industrial Revolution that encourage students to connect past events to modern life.”
Weak: “Write a quiz.”
Strong: “Create a 10-question multiple choice quiz on photosynthesis for GCSE Biology students, with 4 options per question, testing both factual recall and application of concepts.”
2. Provide Context
AI doesn't know your students, your school, or your curriculum. The context you provide shapes the relevance and usefulness of the response.
Include details like: grade level, subject area, student abilities, learning objectives, time constraints, available resources, and any specific requirements or restrictions.
3. Define the Format
Tell AI exactly how you want the output structured. Do you want bullet points, a table, a narrative, or step-by-step instructions? Be explicit.
“Present this as a table with columns for: Learning Objective, Activity, Duration, and Resources Needed.”
4. Set Constraints
Boundaries improve output quality. Tell AI what NOT to include, length limits, vocabulary requirements, or complexity levels.
“Keep explanations under 100 words each. Avoid technical jargon. Do not include any examples involving violence or sensitive topics.”
5. Iterate and Refine
Your first prompt rarely produces perfect results. Use follow-up prompts to refine, expand, or redirect the output.
“This is good, but make the questions more challenging for advanced students.”
“Can you add scaffolding hints for each question?”
“Remove questions 3 and 7, and add two more application-based questions.”
The CRAFT Framework
Use this framework to structure effective prompts:
Context
Who you are, what you're working on, and relevant background
Role
What role should the AI take? (expert, critic, collaborator)
Action
What specific task do you want completed?
Format
How should the output be structured?
Tone
What style or voice is appropriate?
Putting It Together: An Example
Before: Weak Prompt
“Help me teach fractions.”
After: Strong Prompt Using CRAFT
“I'm a Year 5 teacher preparing a maths lesson on adding fractions with different denominators. My class of 28 students has mixed abilities—some are still struggling with equivalent fractions while others are ready for more challenge.
Act as an experienced primary maths specialist. Help me design three differentiated activities that all address the same learning objective but at different complexity levels.
Format this as a table with columns for: Activity Name, Difficulty Level (Foundation/Core/Extension), Description, Materials Needed, and Time Required.
Keep explanations practical and jargon-free—I'll need to explain these to a teaching assistant.”
Key Takeaways
- Specificity beats cleverness—clear prompts outperform creative ones
- Context is king—AI can't read your mind or know your situation
- Format your requests—structure in, structure out
- Constraints improve quality—boundaries focus the output
- Iteration is normal—refine through conversation
Interactive Lab
CRAFT Your Prompt
Build a complete, well-structured prompt using the CRAFT framework
Context
What background information does the AI need?
Include your role (student, intern), the assignment context, what tools you have access to, and any constraints.