Back to Guides
22 March 2026

Claude Code - not as scary as you think

beginnertoolstechnical

The "I've Never Touched a Terminal" Guide you didn’t know you needed

Claude Code: The "I've Never Touched a Terminal" Guide

What it is, why it matters, and how to stop being scared of it.


First Things First: Breathe

I know. You saw the word "Code" and your brain went "nope."

You're imagining a black screen with green text scrolling like you're in The Matrix. You're thinking this isn't for you. You're wrong.

Claude Code is just Claude - the same Claude you already talk to - but instead of living in a pretty chat window, it lives in your terminal (that black or blue screen thing). And because it lives there, it can actually do things on your computer. Not just talk about doing things. Actually do them.

That's the whole difference. Chat Claude talks. Code Claude acts.

Real talk: If you've been using Claude.ai for conversations, personas, and companionship - you already know how to "use" Claude Code. You type. It responds. The only difference is where you type and what it can touch.


What Actually Is Claude Code?

Let's strip away the tech jargon. Here's what Claude Code actually is:

Claude.ai (what you use now):

  • You type in a chat box in your browser or app
  • Claude responds with text, creates documents, searches the web
  • Claude cannot touch your files, your computer, or anything outside that chat window
  • Think of it as talking through a glass wall—Claude can see and hear you, but can't reach through

Claude Code:

  • You type in your terminal (a text-based window on your computer)
  • Claude can read your files, create new files, edit existing ones, run programs
  • Claude can interact with your computer directly—with your permission
  • Think of it as opening the door and letting Claude into the room

The simplest way to think about it: Claude.ai is a phone call. Claude Code is Claude sitting next to you at your desk, hands on the keyboard.


Why Should You Care?

You're not a developer. You don't write code for a living. So why would you want this?

Because Claude Code isn't just for coding. That's the biggest misconception. Here's what it can do for normal humans:

For Your AI Relationship

  • Persistent context: Claude Code remembers the project you're working in. Open it tomorrow and it still knows what's there. And let’s be honest…Opus 1m context window is INSANE
  • MCP connections: Claude Code can connect to external tools - your calendar, your email, your files - giving your AI companion actual access to your world.

For Your Daily Life

  • File organisation: Point Claude at 500 messy PDFs and say "rename these by date and sort them into folders." Done.
  • Data cleanup: Got a spreadsheet that's a disaster? Claude Code can clean it, format it, fix the formulas.
  • Writing projects: Claude can create Word documents, presentations, and formatted files directly—no copy-paste.
  • Research compilation: Give Claude access to a folder of notes and ask it to synthesise a report.

For Building Things (Yes, Even You)

  • Websites and apps: Describe what you want in plain English. Claude builds it. You don't need to understand the code.
  • Automations: Repetitive tasks you do every week? Claude can write scripts to handle them.
  • MCP servers: Want your AI to connect to something specific? Claude Code can build that bridge.
  • Apps: So, so many apps. No idea is “silly”. I thought adding gifs to Hearthline is silly but here we are!

The point: You don't need to understand code to use Claude Code. You need to understand what you want. Claude handles the rest.


The Terminal Is Not Scary (I Promise)

I'm going to be honest with you. The terminal looks like absolute shit compared to a nice chat interface. There's no pretty colours. No buttons. No "send" arrow. Just a cursor blinking at you, waiting.

And that's actually the beauty of it.

The terminal is the most direct way to talk to your computer. No menus to navigate, no buttons to find, no "where the hell is that setting" moments. You type what you want. It happens.

What You're Actually Looking At

When you open a terminal, you'll see something like this:

PS C:\Users\Marta>

That's it. That's the whole thing. It's just telling you - I AM IN THIS PARTICULAR FOLDER

Once Claude Code is running, you won't even see most of that. You'll see Claude's interface - which is actually quite friendly. It uses colour coding, shows you what it's doing, and asks permission before touching anything.

Think of the terminal like texting. You wouldn't say texting is scary just because there's no fancy graphics. You type, they respond. Same energy.


Getting Started: The Light Walkthrough

I'm not going to give you a full installation manual here - Anthropic's docs do that better than I can, and they keep it updated. What I'm going to do is give you the overview so you know what to expect and aren't blindsided.

What You Need

  • A Claude subscription: Pro ($20/month) works. Max ($100/month) gives you way more usage for heavy Claude Code sessions.
  • A computer: Mac, Windows, or Linux. All work.
  • About 15 minutes: Seriously, that's it for basic setup.

The Installation Overview

There are several ways to install Claude Code. Here's the simplest path for each system:

On Mac

Open Terminal (it's in Applications > Utilities, or just search for it with Spotlight). Then type:

brew install claude-code

If you don't have Homebrew installed, you can also use: curl -fsSL https://storage.googleapis.com/anthropic-sdk/claude-code/install.sh | sh

On Windows

Open PowerShell (search for it in Start menu). Then type:

winget install Anthropic.ClaudeCode

You'll need Git for Windows installed first. If you don't have it, download it from git-scm.com.

On Linux

Open your terminal and run:

curl -fsSL https://storage.googleapis.com/anthropic-sdk/claude-code/install.sh | sh

First Launch

After installing, navigate to any folder and type:

claude

That's it. One word. Claude will start up, ask you to log in with your Anthropic account (it opens a browser window for this - easy), and then you're in.

image.png

You'll see a prompt. Type something. Anything. Just like chat.

You: What files are in this folder?
Claude: Let me check...

I can see the following files:
- notes.txt
- photos/
- budget.xlsx

Would you like me to do anything with these?

That's Claude Code. You talk. It does.

If you get stuck: The official docs are excellent and written for beginners. Visit code.claude.com/docs for the full setup guide. You can also just… ask Claude in your normal chat to walk you through it.


Key Concepts (The Stuff That Actually Matters)

CLAUDE.md - Your Partner's Constitution

We’ve previpusly mention how to set it up here > https://aidhd.co/guide/claude-code-persona/

This is probably the single most useful thing in Claude Code for our community.

A CLAUDE.md file is just a text file you create in your project folder. Claude reads it automatically every time you start a session. It's like pinning instructions to Claude's forehead.

What you can put in it:

  • Your partner's personality, name, tone, mannerisms
  • Relationship context and history
  • Project-specific instructions ("always use British English", "never use emoji")
  • Coding standards if you're building something
  • Literally anything you'd put in a system prompt

Why it matters: No more copy-pasting your persona instructions at the start of every conversation. Write it once, save it, and Claude just… knows.

Permissions - Claude Asks Before Acting

Claude Code won't just rampage through your files. Before it creates, edits, or runs anything, it asks you:

Claude wants to create file: index.html
Allow? 

>yes
 no

You're always in control. Nothing happens without your say-so. You can also set it to auto-approve certain actions once you're comfortable.

Context - Claude Sees Your Whole Project

In regular chat, Claude only knows what you paste into the conversation. In Claude Code, it can see your entire project folder -every file, every subfolder. This means you can say things like:

"Read all the files in the /notes folder and give me a summary"

And it just… does it. No uploading. No pasting. It's already there.

MCP (Model Context Protocol) - Claude's Hands

MCP is what lets Claude Code connect to external things—your Google Calendar, GitHub, email, smart home devices, and more. Think of it as giving Claude extra arms.

Setting up MCP is its own topic (and we've covered it in a separate guide), but know that this is where the real magic happens. This is how Claude goes from "smart chatbot" to "actual assistant who does things in the real world."


Not Ready for the Terminal? Try the VS Code Extension

If the terminal still makes you want to hide under a blanket, there's a middle ground.

VS Code is a free code editor from Microsoft. It looks like a normal app with menus and buttons and colours. And Claude Code has an extension for it.

What this gives you:

  • Claude Code in a visual interface (not a scary black box)
  • You can see your files in a sidebar
  • Claude shows file changes as visual diffs (green for added, red for removed)
  • It feels much more like a "normal" app

How to get it:

  1. Download VS Code from code.visualstudio.com (free)
  2. Open VS Code, press Ctrl+Shift+X (or Cmd+Shift+X on Mac)
  3. Search for "Claude Code" and click Install
  4. Open the Command Palette (Ctrl+Shift+P) and type "Claude Code"

My honest take: VS Code is a great stepping stone. Use it to get comfortable with Claude Code's capabilities, then graduate to the terminal when you're ready. Or don't. Both work.


What Can Go Wrong (And Why It's Fine)

Let's address the fears:

"What if Claude deletes my files?"

It asks permission first. Always. And your files aren't gone-gone -they're in your recycle bin. Plus, if you're using Git (which Claude can set up for you), everything is versioned and recoverable.

"What if I break something?"

You're working in a folder you choose. Claude can't access anything outside that folder unless you tell it to. Your system files, your photos, your bank details -all untouched. Create a test folder, play in there.

"What if I type the wrong command?"

Then you'll get an error message. That's it. The computer won't explode. Claude won't judge you. Type the right thing next time. Errors are just the computer saying "I didn't understand that."

"What if I use up all my tokens?"

Claude Code uses your subscription's token allowance. On Pro, you'll hit limits faster with heavy use. On Max, you get significantly more. If you hit the limit, you just… wait. Nothing breaks. You can pick up where you left off.

The golden rule: Start with a test folder. Put some junk files in it. Ask Claude to organise them, rename them, create something new. Break things on purpose. Get comfortable. Then move to real projects.


Tips From Someone Who Was Terrified Too

  1. Be specific with what you ask. "Make the code better" is useless. "Rename all files in this folder to start with the date in YYYY-MM-DD format" is gold.
  2. Use /clear between different tasks. Claude Code builds up context during a session. If you switch from organising files to writing a document, the old context can muddy things up. Fresh start keeps it sharp.
  3. Don't micromanage. Tell Claude what you want to achieve, not how to achieve it. Let it plan. It's surprisingly good at breaking down complex tasks.
  4. Start small. Don't jump straight into "build me a website." Start with "create a text file that says hello." Then "organise these five files." Build confidence.
  5. Read the permission prompts. Don't just mash Y. Read what Claude wants to do. It takes two seconds and saves you from surprises.
  6. Your CLAUDE.md is your best friend. Invest time in writing a good one. It's the single biggest return on time you'll get with Claude Code.

The Honest Conclusion

Claude Code isn't perfect. The terminal isn't pretty. There's a learning curve. You'll feel lost at first and that's completely normal.

But here's what I know: every single person in our community who's tried Claude Code has come back and said some version of "why didn't I do this sooner."

Not because they became programmers. But because they unlocked what Claude can actually do when you take the glass wall away.

Your partner gets a permanent personality file. Your files get organised. Your ideas get built into real things. And you did it yourself—one typed command at a time.

The terminal is just a different door to the same Claude you already know. Walk through it.


Resources:

Comments

Loading comments...