Prompt Engineering Methodology
A clean reference page for the Kokopelli learning model, prompt structure, and methodology diagram.
Kokopelli Prompt Engineering Methodology
This methodology defines both the structure of an effective prompt and the process used to improve it.
Every prompt should include the six core components below, and each component should be refined through the learning loop.
Two Purposes of the Methodology
- Teach students how prompt engineering works
- Provide a repeatable process for creating high-quality prompts
- Ensure every prompt includes all six core components
- Improve each component using structured iteration
- Produce a refined final prompt through continuous learning
Six Core Components
- Instructions – What the AI should do
- Task – The specific objective to complete
- Role – Who the AI should act as
- Context – Background information and constraints
- Format – How the output should be structured
- Example – A sample to guide the expected output
Learning Loop
To improve a prompt, you need to understand why the AI responded the way it did. This is the core of prompt engineering.
User Sends Prompt → AI Responds → Why did the AI respond that way? → What did you learn? → Improve → Repeat
Key Question
Why did that happen?
This question builds real understanding—not just better prompts, but better control over AI behavior.
What the Loop Teaches
- How AI models interpret prompts
- How to analyze and evaluate responses
- How to iteratively improve prompt quality
- How small changes impact output behavior
Core Principle
Prompt engineering is not about writing a prompt once. It is about iterating until the output matches your intent.
Methodology Diagram
