# Kokopelli Prompt Engineering Methodology

## COMPREHENSIVE PROMPT ENGINEERING METHODOLOGY

All prompts must contain the following six components.  
Each component is developed using the learning loop.

1. **TASK**: prompt → output → why_did_that_happen → learn_something → try_again  
2. **ROLE**: prompt → output → why_did_that_happen → learn_something → try_again  
3. **CONTEXT**: prompt → output → why_did_that_happen → learn_something → try_again  
4. **FORMAT**: prompt → output → why_did_that_happen → learn_something → try_again  
5. **EXAMPLE**: prompt → output → why_did_that_happen → learn_something → try_again  
6. **INSTRUCTIONS**: prompt → output → why_did_that_happen → learn_something → try_again  

---

## COMPREHENSIVE PROMPT ENGINEERING METHODOLOGY
![Kokopelli Prompt Engineering Methodology Diagram](kokopelli_methodology_diagram_v2.png)

---

# SUMMARY OF PROMPT ENGINEERING METHODOLOGY

## METHOD

Create prompt components:

1 TASK  
2 ROLE  
3 CONTEXT  
4 FORMAT  
5 EXAMPLE  
6 INSTRUCTIONS  

Apply the learning loop to each component:

prompt → output → why_did_that_happen → learn_something → try_again

---

# HOW TO LEARN THIS METHODOLOGY IN KOKOPELLI

Students learn this methodology through **five exercises**.

In each exercise, the student:

1. Builds a prompt containing the six components.  
2. Improves each component using the learning loop.  
3. Produces a final prompt that solves the exercise task.

---

# PROMPT ENGINEERING COMPONENTS

TASK  
ROLE  
CONTEXT  
FORMAT  
EXAMPLE  
INSTRUCTIONS  

These six components form the **structure of a well-designed prompt**.

---

# LEARNING LOOP

prompt → output → why_did_that_happen → learn_something → try_again

This loop is used to improve each prompt component.

The loop teaches students:

- how AI models respond to prompts  
- how to analyze AI outputs  
- how to iteratively improve prompts  

---

# TWO PURPOSES OF THE METHODOLOGY

1. It teaches students **how prompt engineering works**.  
2. It provides a **practical process they can use anytime they create a prompt**.

---

# INSTRUCTIONS FOR STUDENTS

Students **do not delete earlier sections of the prompt**.

Instead, they:

- add each component step by step  
- refine each component using the learning loop  
- continue until the final prompt is complete