Prompt Engineering for Developers

Reliable patterns, structured outputs and tool use

By Houssam Kodad

PDF 144 pages Beginner English

One-time purchase

€19.95

VAT included
where applicable

Download sample

Instant download after purchase
Readable on any device
Free updates to this edition
Secure checkout

About this book

What's inside

Prompt engineering is less about clever wording and more about disciplined software design. This concise, example-driven book teaches developers the patterns that make LLM outputs reliable enough to build on: clear instructions, structured outputs, decomposition and tool use. You'll leave able to write prompts you can test, version and trust inside a real application.

What you'll learn

Skills you'll walk away with

Write clear, testable instructions for an LLM
Use roles, delimiters and examples effectively
Get structured JSON outputs you can parse safely
Decompose hard tasks into reliable steps
Apply chain-of-thought and self-checking patterns
Define tools and function calls cleanly
Version, test and iterate on prompts like code

Table of contents

8 chapters

01
Prompting as Software Design
- · Why prompts are interfaces
- · Determinism and variance
- · Treating prompts as code
02
Anatomy of a Good Prompt
- · Instructions and constraints
- · Roles and delimiters
- · Examples that help
03
Structured Outputs
- · Asking for JSON
- · Schemas and validation
- · Recovering from bad output
04
Decomposition and Chaining
- · Breaking down tasks
- · Prompt chains
- · Routing between prompts
05
Reasoning Patterns
- · Chain-of-thought
- · Self-consistency and checking
- · When reasoning hurts
06
Tool Use and Function Calling
- · Describing tools clearly
- · Argument validation
- · Handling tool errors
07
Reducing Hallucination
- · Grounding with context
- · Asking the model to abstain
- · Citations and verification
08
Testing and Iterating on Prompts
- · Small evaluation sets
- · Versioning and diffs
- · Measuring improvement