Skip to main content
โ† Back to Vibe Coding samples
๐Ÿค–Vibe Codingยท15 minยทSample Lesson

From Plain-English Prompts to Working Apps

A student wants a quiz app to study for a biology test. Instead of memorizing code syntax, she types: "Build a web page that shows one flashcard at a time with a Next button." A few seconds later, an AI coding assistant hands back working code. She tests it, asks for a Shuffle button, and ships it that afternoon. That is vibe coding: building real software by describing what you want in plain language and steering an AI to get there.

What You'll Learn

By the end of this lesson you will be able to: 1. Define vibe coding and how it differs from traditional coding. 2. Write a clear, specific prompt that gets useful code. 3. Use a build-test-refine loop to improve an app step by step.

What Vibe Coding Really Means

In traditional coding, you write every line of syntax yourself, following strict grammar rules. One missing semicolon can break everything. In VIBE CODING, you describe your goal in natural language and an AI model writes the code. Your job shifts from typing syntax to being the director: you decide what to build, judge whether the result is good, and ask for changes. You still need to understand what the code does, but you spend more time thinking about the idea and less time fighting punctuation.

Writing a Prompt That Works

The quality of your code depends on the quality of your prompt. Vague prompts get vague results. Weak prompt: "Make a game." Strong prompt: "Make a browser game where the player moves a green square with arrow keys to catch falling circles. Add a score counter in the top corner and speed up every 10 points." The strong prompt names the controls, the goal, the look, and the rules. The AI has enough detail to build something close to what you actually pictured.

AI can be confidently wrong

An AI assistant will sometimes hand you code that looks perfect but does not run, or that solves the wrong problem. Never trust code you have not tested. Running it and reading the error messages is part of your job, not the AI's.

The Build-Test-Refine Loop

Good vibe coders work in a loop: 1. BUILD: describe a small feature and let the AI generate it. 2. TEST: run it and see what actually happens. 3. REFINE: tell the AI what is wrong or what to add, then repeat. You do not describe the whole app in one giant prompt. You grow it one feature at a time, checking after each step. This is faster and easier to fix than trying to build everything at once.

Match each step of the loop to what you do.

Terms

Build
Test
Refine
Prompt

Definitions

Describe a feature and generate code
The plain-language request you write
Run it and watch what happens
Tell the AI what to fix, then repeat

Drag terms onto their definitions, or click a term then click a definition to match.

โ“

Why is a specific prompt usually better than a vague one?

Breaking a Big Idea into Small Prompts

Say you want a budgeting app. Do not ask for "a budgeting app" all at once. Break it down: - Prompt 1: a form to type in an expense and its amount. - Prompt 2: a list that shows all expenses added so far. - Prompt 3: a total at the bottom that adds them up. - Prompt 4: a button to delete an expense. Each small prompt is easy to test. If step 3 breaks, you know exactly where the problem is.

โ“

What is the main advantage of breaking an app into several small prompts?

๐ŸŽฏ

Plan an App in Five Prompts

Pick a simple app you would like to exist (a habit tracker, a tip calculator, a trivia game). Write out five prompts in order that would build it one feature at a time, starting with the most basic version and adding one thing per prompt. For each prompt, write one sentence describing how you would TEST that the feature works. Trade your prompt plan with a classmate and see if they can spot any step that is too big and should be split.

Want to keep learning?

Sign up for free to access the full curriculum โ€” all subjects, all ages.

Start Learning Free