Install Claude Code and run your first session

Claude Code is a coding assistant that runs in your terminal. Unlike a chat window, it works inside your project folder. It can read files, search them, run commands, and propose edits, asking your permission along the way. This club is independent and not affiliated with Anthropic, so treat the official docs at code.claude.com/docs as the source of truth. Install steps change, and this lesson reflects the docs as we last checked.

What you need

• A terminal (Terminal or iTerm on Mac, a shell on Linux, PowerShell or WSL on Windows).
• A paid Claude account or a Console account for API billing. The free claude.ai plan does not include Claude Code, according to the setup docs. Plans change, so confirm on the current page.
• A project folder. Any small folder works, ideally one under git.

Install

The setup page lists several methods. The native installer on macOS, Linux, and WSL is:

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

On Windows PowerShell:

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

Homebrew (brew install --cask claude-code) and WinGet (winget install Anthropic.ClaudeCode) are also listed, as is an npm package that needs a recent Node version. Pick one method and stick with it.

Piping a script from the internet into your shell is a trust decision. Read the official page, make sure the URL matches, and if you prefer, download the script first and read it.

Check that it worked:

claude --version
claude doctor

claude doctor runs read-only diagnostics and tells you if something is off.

Start a session and sign in

Go to your project folder and run:

cd ~/projects/my-site
claude

The first run opens a browser so you can sign in. After that you land in an interactive prompt. Anything you type is a request. Type /help to see commands.

Your first task: read-only

For a first session, ask it to look, not change:

Give me a tour of this project. What does it do,
where is the entry point, and how do I run it?

Claude will read files and may ask permission for some actions. Read each prompt before approving. Saying no is always allowed, and you can ask why it wanted to do something.

Habits to start now

• Start from the right folder. Claude Code works relative to where you launched it.
• Use /clear when you switch to an unrelated task, so old context does not leak into the new one.
• Use /resume to return to an earlier session.
• Use /model to see or change the model for the session.
• Run it in a git repo. Git gives you a safety net, which lesson 5 covers.

When something goes wrong

If the command is not found, close and reopen the terminal so your path updates, then run claude doctor. If sign-in fails, re-check that your account type includes Claude Code. The troubleshooting section of the setup docs is the best next stop.

Try it

Install Claude Code, open a small project, and run the tour prompt above. Then ask two follow-ups: "Which file would I change to rename the site title?" and "What would you want to check before changing it?" Do not approve any edits yet. Write down which permission prompts appeared and what each one asked for.