Anthropic recently released an open-source interactive prompt engineering tutorial on GitHub, with a clear structure suitable for learners of different levels.


The tutorial is divided into three main levels: beginner, intermediate, and advanced. The beginner section covers basic prompt structures, clear expression, and role assignment; the intermediate section delves into data and instruction separation, formatting, and step-by-step reasoning; the advanced section focuses on avoiding hallucinations, industry applications, and complex prompt design. The appendix also includes advanced topics such as prompt chains, tool usage, and search strategies.

A developer, Hemachandiran, shared their insights on prompt engineering, proposing four practical tips: keeping requirements clear and specific, using positive instructions, avoiding overly long variable names, and breaking down complex tasks using a chain-of-thought pattern.


The interactive design of this tutorial allows learners to test different prompt strategies in real-time, making it more practical than traditional document-based tutorials. For developers and AI enthusiasts looking to systematically learn prompt engineering, this is a valuable learning resource worth investing time in.

The tutorial is currently open-sourced on GitHub under the MIT license, allowing free use and modification. This open approach contrasts with the generally closed practices of current AI companies, providing substantial help for community learning.