When you hear "let the AI write code," you might picture yourself copy-pasting into a chat window. But today's AI coding tools left that stage behind long ago. They read your files, rewrite your code, run your tests, and fix their own errors―driving an entire project like a partner sitting right beside you. This course guides you, by the fastest route, all the way to actually mastering the major tools: Claude Code, Cursor, GitHub Copilot, and more.

What you'll be able to do after this course

The goal: make the AI your partner and move 10x faster

Choose the right tool
Understand the differences between CLI, editor, and cloud tools, and pick the partner that fits how you work.
Use it the right way
From setup and permission settings to giving instructions, you'll learn a safe, efficient way to work.
Get unstuck when you're stuck
From handling common errors and managing costs to extensions, you'll gain the practical know-how to keep working smoothly.

What are AI coding tools?

"AI coding tools" is the umbrella term for development tools that use large language models (LLMs) to help you write, fix, and debug code. They range widely, from simple "smart autocomplete" to "autonomous agents" that carry out an implementation across multiple files from a single instruction.

What they all share is this: they turn the intent you express in plain language into working code. Work that used to mean hours of digging through documentation and trial and error can now be done in one go, through conversation. That said――

💡 You're still the one in charge. The AI won't decide what to build. Giving precise instructions, reading the code that comes back, and judging whether it's good is the human's job. Become someone who truly commands the tool and your productivity soars; hand everything off blindly and you pile up code that looks like it works but is dangerously fragile. This course is your map to becoming the one in command.

Three types ― CLI, editor, cloud

However many tools there are, sorting them by where they run into three broad types makes the whole landscape snap into focus. Fix this map in your mind first.

① CLI agents
Autonomous, running in the terminal

Give an instruction from the terminal and the AI reads and writes files itself, runs commands, and even runs your tests. You can hand it an entire project. The flagship is Claude Code.

Best for: people who want to delegate large chunks of implementation.

② AI editors
A partner built into your editor

The AI handles completion, chat, and edits right inside the code editor you already know. The appeal is the reassurance of working in dialogue while watching the code. Cursor and GitHub Copilot are the leading examples.

Best for: people who want to check the code as they write.

③ Cloud app builders
Done entirely in the browser

No environment setup needed: describe what you want in the browser and it generates the whole foundation of an app. Tools like v0, Bolt, and Lovable. You can get something real with almost no code.

Best for: people who want to see something working quickly first.

These three types are not mutually exclusive. Building the foundation with an app builder, polishing it in an AI editor, then adding features with a CLI agent is a perfectly normal combination. This course focuses mainly on ① CLI agents (Claude Code) and ② AI editors, digging deep into using them in real work. Type ③ is covered in detail in the AI app builder comparison article and in the companion course "Solo development with AI."

The major tools at a glance

Here are the major tools worth knowing right now, laid out by type and characteristics. Grasp their names and where each one stands, and the chapters ahead will read much more easily.

Claude Code CLI agent

Made by Anthropic. Runs in the terminal and handles planning, implementation, and testing autonomously. The star of this course. Official site: claude.com/claude-code.

Cursor AI editor

An AI-focused editor built on VS Code. Talk with the AI and make bulk edits while watching the code. Learn more in What is Cursor.

GitHub Copilot AI editor

The most widely adopted completion-style tool. Built into VS Code and others, it suggests the next line as you type. Learn more in What is GitHub Copilot.

Codex CLI/cloud

OpenAI's coding agent. It sits in a position close to Claude Code. See how the two differ in Claude Code vs Codex.

v0 / Bolt / Lovable App builders

A family of near-no-code tools that generate an app from instructions you give in the browser. See the app builder comparison.

📊 If you want to compare all four tools side by side, the in-depth comparison of Cursor, Claude Code, GitHub Copilot, and Codex is handy. Among them, this course centers its hands-on work on Claude Code, which can take on large parts of the implementation.

Which one should you pick?

"So which one should I actually use?"――that's the first hurdle. The answer comes down to how you work. Try answering these three questions.

Q1. Can you read code?

Barely at all → start with an app builder or vibe coding. A little → move to an AI editor or Claude Code.

Q2. How much do you want to hand off?

Delegate implementation in one go → Claude Code (CLI agent). Confirm line by line → Copilot or Cursor.

Q3. Do you want to set up an environment?

Rather avoid it, want to try right now → a browser-based app builder. Work seriously in your own environment → CLI/editor tools.

✅ Not sure? Start with Claude Code. It can take on large parts of the implementation, and all you need to begin is a terminal. This course also builds around Claude Code, so the most direct route is to get comfortable with it first and add Cursor or Copilot as needed.

A map of this course

This course has seven chapters in all. In the order "know → begin → compare → ask well → get unstuck → extend → keep costs down," it will get you using these tools at a professional level. Use it as a map of where you are in your learning.

CHAPTER 2
Getting started with Claude Code

Setup, first steps, permission modes, and settings basics.

Chapter 2 →
CHAPTER 3
Comparing tools and using them for the right job

The differences between Cursor, Copilot, and Codex, and how to choose.

Chapter 3 →
CHAPTER 4
Asking well

When to use vibe coding versus spec-driven development.

Chapter 4 →
CHAPTER 5
Getting unstuck

The causes of common errors and how to handle them, laid out systematically.

Chapter 5 →
CHAPTER 6
Using extensions

Expand your capabilities with hooks, plugins, subagents, and MCP.

Chapter 6 →
CHAPTER 7
Cost and efficiency

Managing tokens and usage limits, plus tips for working efficiently.

Chapter 7 →

Before you begin

From the next chapter on, you'll actually get your hands moving. There isn't much to prepare. Being a little comfortable with the terminal, and having some code you want to build (practice code is fine)―that's all you need.

🖥 Terminal basics

A rough sense of changing folders and running commands is enough. It doesn't have to be perfect.

📁 A practice project

Have one thing you're free to tinker with, like a small to-do app or a script.

🔑 An account

An account for the tool you'll use (Anthropic, for Claude Code). Details in the next chapter.

In this chapter
  • AI coding tools are partners that turn plain-language intent into working code. You're still the one in charge.
  • There are three types ― ① CLI agents (Claude Code), ② AI editors (Cursor/Copilot), ③ cloud app builders (v0/Bolt/Lovable). You can mix them.
  • How to choose comes down to "Can you read code? / How much do you want to hand off? / Do you want to set up an environment?" When in doubt, start with Claude Code.
  • This seven-chapter course centers on Claude Code and guides you all the way to professional-level mastery.

Let's get started. In the next chapter, Chapter 2, "Getting started with Claude Code," you'll actually install it and go all the way to giving your first instruction.