Part II: Core Frameworks

Architectural Patterns and Design Principles for Effective Prompts

Overview

Part II dives deep into the structural elements of prompt design. Once you understand the foundations, you need practical frameworks for constructing prompts that work. This section provides architectural patterns, context engineering strategies, instruction design principles, and output specification techniques.

These chapters cover:

How to structure prompts for maximum effectiveness
Techniques for providing the right context
Methods for writing clear, actionable instructions
Ways to specify exactly what output you need

Chapters in This Part

Chapter 4: Prompt Architecture Patterns

Structural frameworks for prompt design, including:

The anatomy of an effective prompt
Common architectural patterns (CRAFT, RISEN, etc.)
When to use which pattern
Building blocks and composability

Chapter 5: Context Engineering

Mastering the art of context, including:

What context is and why it matters
Types of context (domain, task, user, system)
Context loading strategies
Managing context window constraints

Chapter 6: Instruction Design Principles

Writing effective instructions, including:

The instruction hierarchy
Clarity and specificity techniques
Handling ambiguity
Instruction sequencing and priority

Chapter 7: Output Specification Techniques

Controlling AI outputs, including:

Format specification methods
Structured output (JSON, XML, Markdown)
Length and detail control
Quality constraints and validation

Key Frameworks Introduced

Framework	Purpose
CRAFT	Context, Role, Action, Format, Tone
RISEN	Role, Instructions, Steps, End goal, Narrowing
CO-STAR	Context, Objective, Style, Tone, Audience, Response
RTF	Role, Task, Format

The Prompt Architecture Diagram

┌─────────────────────────────────────────────────┐
│                   PROMPT                        │
├─────────────────────────────────────────────────┤
│  ┌─────────────────────────────────────────┐   │
│  │           CONTEXT LAYER                 │   │
│  │  Background │ Domain │ Constraints      │   │
│  └─────────────────────────────────────────┘   │
│  ┌─────────────────────────────────────────┐   │
│  │         INSTRUCTION LAYER               │   │
│  │  Role │ Task │ Requirements │ Steps     │   │
│  └─────────────────────────────────────────┘   │
│  ┌─────────────────────────────────────────┐   │
│  │           OUTPUT LAYER                  │   │
│  │  Format │ Structure │ Constraints       │   │
│  └─────────────────────────────────────────┘   │
│  ┌─────────────────────────────────────────┐   │
│  │          EXAMPLES LAYER                 │   │
│  │  Few-shot examples │ Counter-examples   │   │
│  └─────────────────────────────────────────┘   │
└─────────────────────────────────────────────────┘

Learning Objectives

After completing Part II, you will be able to:

Select appropriate prompt architecture patterns for different tasks
Engineer context effectively within token constraints
Write clear, unambiguous instructions
Specify output formats that meet your needs
Combine these elements into well-structured prompts

Practical Exercises

Each chapter in Part II includes hands-on exercises:

Pattern application exercises
Context optimization challenges
Instruction refinement practice
Output specification drills

Prerequisites

Completion of Part I: Foundations, or equivalent understanding of:

Basic prompting concepts
The 7 Prompting Pillars
Token and context window basics

Estimated Reading Time

Chapter 4: 25-30 minutes
Chapter 5: 30-35 minutes
Chapter 6: 25-30 minutes
Chapter 7: 25-30 minutes

Total: Approximately 2-2.5 hours

Next Steps

Begin with Chapter 4: Prompt Architecture Patterns to learn the structural foundations of effective prompt design.

Begin Part II →