Welcome
You do not need to memorise this guide. Use it to recognise the language, understand the workflow, and know what to ask next.
What this guide is
Vibe coding means creating software by describing what you want, reviewing what the AI produces, and guiding it through a series of improvements. You are not required to write every line of code yourself. You are required to think clearly, make decisions, and recognise when the result is not good enough.
That makes vibe coding surprisingly familiar to designers and researchers. We already know how to frame ambiguous problems, understand people, establish principles, create systems, critique work, and iterate. The interface has changed — from a canvas or document to a conversation with an agent — but the need for judgement has not.
This guide gives you enough technical understanding to participate confidently without pretending that a short course turns everyone into a software engineer. It focuses on the parts that transfer immediately into your work: framing, context, collaboration, prototyping, validation, and clear communication.
How to use it
Read the guide once in order. The chapters build on one another. During practical work, return to the templates, checklists, and command reference rather than trying to remember everything.
What you will be able to do
By the end, you should be able to:
- explain what vibe coding is without reducing it to “prompt and hope”;
- choose an appropriate tool for a small prototype;
- give an AI enough context to make useful decisions;
- break an idea into small, testable slices;
- contribute through your own discipline rather than imitating a developer;
- recognise basic code, terminal, Git, API, and deployment language;
- review a generated experience for usability, content, accessibility, responsiveness, and risk;
- build and publish a small interactive webpage with AI assistance.