Prompt Engineering Resources

These resources provide a practical path for learning prompt engineering. Start with the roadmap and general guide, then move into model-specific guidance and context engineering.

Prompt Engineering Roadmap

Why this is the starting point

If you want a direction for learning prompt engineering for a career in AI, this roadmap gives you small explanations of each topic.

You will need to study each topic in more detail and practise it yourself. Keep the roadmap open as you work through the other guides. It provides the vocabulary and context needed to understand them.

Prompt Engineering Guide

Why this is a good source

This free guide takes you from basic prompting techniques to more complex methods for getting better outputs.

It is my preferred general guide for getting started with prompt engineering. It also explains different models, how they work, their strengths, and how to prompt them effectively.

If you understand the material and practise it, you will have a strong foundation in prompt engineering.

The Ultimate Guide to Prompt Engineering in 2026

For more specific knowledge

This free guide is less detailed than the previous guide, but it provides focused and recent information that you can apply quickly.

Its most useful section explains how to combine prompt types to make prompts more effective.

Prompting best practices by Anthropic

Prompting best practices

Next step: model-specific knowledge

This guide explains how to prompt Claude for useful code and how to control its output.

Each language model has specific differences. For example, Anthropic recommends placing long-form data near the top of a prompt. This can improve performance across Claude models.

The guide assumes you already understand the topics covered by the general resources above.

Best Practices for Claude Code

Personalise your workspace and prompts

This guide explains how to personalise Claude Code, manage its weaknesses, and control its use of context and tokens.

It also explains features such as planning mode and edit mode. Personalising your tools and building a repeatable workflow makes the tool more useful for your specific work.

GPT-5.1 Prompting Guide

OpenAI's model-specific guide

This detailed guide covers techniques for working with GPT-5.1. It assumes that you already understand language models and have experience writing prompts.

Use it to improve how you work with ChatGPT, Codex, and GPT-5.1.

Design strategies for the Gemini API

Gemini's model-specific guide

Use this guide to practise prompt engineering with Gemini. Gemini differs from ChatGPT and Claude, particularly when working with structured and visual material such as PDFs and research papers.

Prompt engineering versus context engineering

Why AI Teams Are Moving From Prompt Engineering to Context Engineering

Move from prompts to systems

Once you are comfortable writing prompts, context engineering is the natural next step.

Prompt engineering alone does not solve memory, retrieval, or multi-step work where the available context changes. Context engineering gives you more control over the information and tools available to a language model. This article is a useful introduction to that shift.