Claude Code for beginners: a tutorial from install to your first project

What is Claude Code?

Claude Code is an agentic coding tool that reads your codebase, edits files, runs commands and integrates with your development tools. The word that matters is agentic. In a chat window, you copy code back and forth. Claude Code works inside your project folder: you describe what you want, and it plans the approach, writes the code across files and checks that it works.¹

It runs on several surfaces that share the same engine, so your CLAUDE.md files and settings work on all of them.¹ Here’s how they compare:

What you need to start

• A Claude subscription (Pro, Max, Team or Enterprise), a Claude Console account, or access through a supported cloud provider.² The Free plan doesn’t include Claude Code.³
• A terminal and a project folder. Never opened a terminal? Anthropic has a terminal guide,² or you skip the terminal and use the desktop app.¹
• On native Windows, Git for Windows is recommended so Claude Code can use the Bash tool.²
• One thing to know about limits: Claude Code shares them with the rest of your plan, so terminal work and chats draw from one pool.³

Install Claude Code with the official commands

These are the native install commands from Anthropic’s quickstart, the method it recommends.² Commands can change, so compare them with the docs before you paste anything into a terminal.

macOS, Linux or WSL:

curl -fsSL https://claude.ai/install.sh | bash

Windows PowerShell:

irm https://claude.ai/install.ps1 | iex

Windows CMD:

curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd

Homebrew (brew install --cask claude-code) and WinGet (winget install Anthropic.ClaudeCode) work too, but unlike the native install they don’t update automatically.² When the installer finishes, open a new terminal window and check that it worked:

claude --version

You should see a version number followed by “(Claude Code)”. Then move into your project folder and start your first session. On first use, Claude Code asks you to log in through your browser.²

cd /path/to/your/project
claude

How to talk to Claude Code

Start with a question, not a change. Anthropic’s quickstart suggests opening with “what does this project do?” Claude reads the files it needs, so you don’t have to paste context.² Then hand out work the way you’d brief a sharp colleague:

• Be specific. Name the file, describe the symptom and say what “fixed” looks like.⁴
• Hand over what you have. Type @ and a file path to include a file, or paste a screenshot.⁴
• Not sure what you want yet? Ask Claude to interview you until it could write the spec, then build from that.⁴
• Give it a check to run, like a test or a script, and ask for evidence: the command it ran and the output.⁴
• Split big jobs into numbered steps.²

Three commands cover most beginner sessions: /help lists what’s available, /clear starts the conversation fresh, and claude -c picks up your most recent conversation in the current folder.²

CLAUDE.md basics

CLAUDE.md is a plain markdown file that Claude Code reads at the start of every session. For a project, it lives in ./CLAUDE.md or ./.claude/CLAUDE.md. You don’t have to start from scratch: /init analyzes your project and writes a first draft with the build commands, test instructions and conventions it finds.⁵

Anthropic’s rule of thumb: write down what you’d otherwise re-explain. Add a line when Claude makes the same mistake twice, keep each file under about 200 lines, and make instructions concrete enough to check, like “Run npm test before committing” instead of “Test your changes”. Claude treats the file as context, not as enforced configuration.⁵ A short example for a video project:

# My reel project

- Videos are vertical, 1080 × 1920
- Scripts run 150 to 170 words for 60 seconds
- Source images live in assets/. Never delete them.
- After every render, tell me the file path
  and the length

We go deeper on rule files and skills in Claude skills and CLAUDE.md.

Safe habits from day one

• Know your permission mode. It decides what Claude may do without asking. In Manual mode, Claude Code stops and asks before most actions that edit files, run shell commands or reach the network. Recent versions start new terminal and VS Code sessions in auto mode, where a second model reviews actions instead of you, and Anthropic notes that auto mode doesn’t guarantee safety. Press Shift+Tab to switch modes, for example to Manual while you learn.⁶
• Plan before bigger changes. In Plan mode, Claude researches and proposes changes but doesn’t edit your source until you approve the plan. Start a session in it with claude --permission-mode plan.⁶
• Review what changed. Ask “what files have I changed?”² or look at the diffs in the desktop app or VS Code.¹
• Use checkpoints. Every prompt creates one, and /rewind, or Esc twice on an empty prompt, restores your code, the conversation or both. Files that only shell commands changed aren’t tracked.⁷
• Keep Git anyway. Checkpoints are for quick recovery inside a session and don’t replace version control.⁷
• Keep secrets out of CLAUDE.md. Project instructions are meant to be shared through version control,⁵ so API keys belong in your own environment, not in that file.
• Leave bypass mode alone. Anthropic reserves bypassPermissions for isolated containers and VMs.⁶

Beginner Claude Code projects

You learn Claude Code by finishing things. Good first projects are small, useful and easy to judge:

1. Explain an old folder. Point Claude Code at a project you’ve forgotten and ask what it does.
2. Build a one-page website. One HTML file, then change it in plain words: bigger headline, darker background, a contact link.
3. Tidy your files. Rename 200 photos by date. Make a copy or commit first, so you can undo.
4. Turn a spreadsheet into a chart. CSV in, image out.
5. Render a short video. Start with a five-second test clip, then build up to a full reel.

Project idea: render a short video

Video is a good teacher because you judge the result in seconds: you press play. A first prompt can be this simple:

make a 5-second vertical test video, 1080 × 1920,
white text "Hello" on a black background.
Save it as test.mp4, then show me
the command you ran and its output.

If it needs a tool you don’t have yet, such as ffmpeg, ask Claude how to install it. From there, a full reel is the same loop with more steps. Our workflow: pick a question, let Claude write a plan with sourced facts, three hook options and a script of roughly 150 to 170 words, have an AI voice read it and return a timestamp for every word, build 3D scenes where each image lands on its word, and render a vertical 1080 × 1920 MP4. Before handover, Claude checks the loudness, keeps text in the safe zone and flags any moment longer than 1.5 seconds without something new on screen.

If you’d like this as a guided project, our course shows you how to make 3D explainer reels with Claude Code. It comes with a guide (PDF) and a Claude kit (ZIP): a CLAUDE.md that Claude Code reads automatically when you open the folder, four rule files (hooks, voice and timing, sync, motion) and a starter project that renders a short sample reel. You need a paid Claude plan with Claude Code and an ElevenLabs account. Explain It in 60 is an independent Get99Skills course, not an Anthropic product.