Beginner~4 weeks6 steps

Prompt Engineering

Get consistently better outputs from AI models: core prompting techniques, reasoning strategies, structured output, evaluation, and prompt security. No coding required to start.

1
How LLMs read your prompt
Tokens, context windows, and sampling — the mental model behind every technique.
- ArticlePrompt Engineering Guide — Introduction
- VideoIntro to Large Language Models (Andrej Karpathy)
2
Core techniques
Clear instructions, roles, examples (few-shot), and output formatting.
- DocsPrompt Engineering Overview (Anthropic docs)
- ArticlePrompt Engineering Guide — Techniques
3
Reasoning techniques
Chain-of-thought, decomposition, and self-consistency for harder tasks.
- DocsChain of Thought Prompting (Anthropic docs)
- ArticleChain-of-Thought (Prompt Engineering Guide)
4
Structured output
Get JSON and machine-readable answers you can build software on.
- ArticleStructured Outputs (OpenAI docs)
5
Evaluating & iterating
Treat prompts like code: test cases, comparisons, and systematic iteration.
- CourseAnthropic Courses — Prompt Evaluations (free, on GitHub)
- ArticleYour AI Product Needs Evals (Hamel Husain)
6
Prompt security
Understand prompt injection and jailbreaks — and how to defend against them.
- ArticleAdversarial Prompting (Prompt Engineering Guide)
- ArticleOWASP Top 10 for LLM Applications

How LLMs read your prompt

Core techniques

Reasoning techniques

Structured output

Evaluating & iterating

Prompt security