You've probably used apps in three different ways. Let's map Claude to those experiences to build a clear mental model.
Click each panel to explore the analogy and see how it relates to Claude.
Tap & Swipe Interface
Click & Type Interface
Connect & Automate
You interact through buttons and screens, and the result appears instantly. The Claude equivalent is the claude.ai website: a visual, chat-based interface where your browser sends a message and gets a response. It's great for conversations and summarizing pasted text.
This is software installed ON your computer that works with files ON your computer (like Microsoft Excel). While there's no direct Claude equivalent, this is the feeling of Claude Code. You type commands, and it makes changes to your local files.
This is when one tool sends a request to an external service for data. A weather widget on your blog is a perfect example. This is what Claude Code is! Your terminal sends a request to Claude's AI brain on Anthropic's servers, gets a response (the code), and then uses that response to change your local files.
Drag each item to the category that best describes its interaction style.
Tap & Swipe
Click & Type
Connect & Automate
Q1: Claude Code is best described as what?
Perfect for small teams: Claude Code runs locally on your computer and connects securely to the Claude API. Whether you're a solo consultant or a 5-person team with a $20/month budget, you get the same powerful capabilities as enterprise teams.
No enterprise licensing required. You pay only for what you use through token-based pricing. A typical small team might spend $10-30/month on actual AI usage, making it accessible even for bootstrapped consultancies.
No hidden costs: Editing sessions inside your IDE are not billed per token directly. You only pay for the AI processing when you actually use it. Organizations can provision access through individual seats or workspace licenses, but even individual consultants can get started immediately with a personal Anthropic account.
Whether you're a solo financial consultant or part of a 50-person firm, you get access to the same Claude AI capabilities. The difference is only in how you manage access and billing, not in the features available to you.
CLAUDE.md to minimize repeated instructionsToday you can ask Claude to edit files with free‑form requests or explicit verbs. Both styles are supported and may converge over time:
claude "edit this file: add a button to index.html"claude edit index.html "add a button"| Command | What It Does | Example |
|---|
// Before
<h1>Hello</h1>
// After
<h1>Hello</h1>
+ <button id="cta">Click Me</button>
⏱️ Time needed: 10-15 minutes | 📚 Experience required: None — we'll guide you through every click!
What is a Terminal? Think of it as a text-based way to talk to your computer. Instead of clicking icons, you type commands. Don't worry — we'll tell you exactly what to type!
🍎 On Mac:
🪟 On Windows:
✅ You'll know it worked when: You see a window with text and a blinking cursor waiting for you to type.
What is Node.js? It's free software that helps run modern programs like Claude Code. Think of it as the engine that makes Claude Code work. You need this first.
📝 Copy this command to check if you already have it:
node --versionPaste it in your Terminal/PowerShell and press Enter.
✅ If you see something like:
v18.17.0, or a similar numberGreat! You have Node.js. Skip to Step 3 below.
⚠️ If you see:
command not foundYou need to install Node.js first. Follow the instructions below.
node --version again to confirm it worked — you should now see something like v22.20.0💡 Which version? The green button automatically gives you the recommended "LTS" (Long Term Support) version, which is perfect for your needs. Don't worry about version numbers — any recent version works fine with Claude Code.
What you're doing: Now that Node.js is installed, you can download and install Claude Code, just like installing Microsoft Word or Excel.
📝 Copy this command exactly as written:
npm install -g @anthropic-ai/claude-codeNow do this:
The Terminal will show several messages. Here's what they mean:
changed 2 packages in 18s✅ This is success! The installation worked. You might see "added 2 packages" or "changed 2 packages" — both mean the same thing. The numbers and time might be slightly different — that's completely fine.
1 package is looking for funding
run `npm fund` for details✅ Ignore this completely. It's just a message from the open-source community asking for donations. You don't need to do anything. Claude Code is already installed and ready to use.
found 0 vulnerabilities✅ Nice! If you see this message, it confirms the software is safe and secure. But don't worry if you don't see it — it doesn't always appear, and that's normal.
💡 The key message: As long as you see "changed 2 packages" (or "added 2 packages") and your cursor is blinking again with no red "ERROR" text, you're done! Everything else is just informational noise you can ignore.
Now that you've successfully installed Claude Code, let's decode what those messages actually meant. Understanding this will make you more confident next time you work with similar tools.
Think of a package as a pre-built software component — like a Lego brick. Instead of building everything from scratch, developers download and use packages that others have already created and tested.
Real-world analogy: It's like buying a pre-assembled cabinet from IKEA instead of cutting wood and building from scratch. Claude Code is a package, and it might depend on other smaller packages to work.
npm stands for "Node Package Manager." It's like the App Store for developers. When you typed npm install, you told npm: "Go to the store, find Claude Code, and install it on my computer."
Real-world analogy: npm is like Amazon for software. You tell it what you want, and it downloads and sets it up for you. The difference? Almost everything on npm is free and open-source!
Most packages on npm are created by volunteers in their free time and shared for free. "npm fund" is like a tip jar — it shows you which packages accept donations to support the developers who maintain them.
Real-world analogy: It's like a busker (street musician) playing beautiful music for free, with a hat on the ground where you can drop a coin if you want. You're not required to donate — the music (software) is free either way.
A vulnerability is a security weakness in software that hackers could potentially exploit. When npm says "found 0 vulnerabilities," it means it scanned the packages and didn't find any known security issues.
Real-world analogy: It's like a home security inspection. "0 vulnerabilities" means all the locks work and there are no broken windows. If it finds issues, it will tell you how to fix them.
The -g flag stands for "global." It means "install this package so I can use it anywhere on my computer, not just in one folder." That's why you can type claude from any location in your Terminal.
Real-world analogy: It's like installing Microsoft Word on your computer (global) versus saving a document to your Desktop (local). Word is available everywhere; the document is only in that one spot.
💪 Why this matters: You just learned the language of modern software development! Next time you see npm, packages, or terminal commands, you'll understand what's happening under the hood. You're not just following instructions — you're becoming tech-literate.
✅ Success looks like: You see messages like "added 2 packages" and your cursor is blinking again, waiting for your next command. No red "ERROR" text.
Why? Your computer needs to "refresh" to recognize the new program you just installed. Think of it like restarting after a software update.
How to do it:
What you're doing: Connecting your computer to your Claude account, like logging into Gmail or Netflix.
📝 Copy this command:
claude auth loginThen:
You'll see an orange/yellow box with text that says:
Do you trust the files in this folder? C:\Users\[YourName] Claude Code may read, write, or execute files contained in this directory. This can pose security risks, so only use files from trusted sources. Learn more ( https://docs.claude.com/s/claude-code-security ) > 1. Yes, proceed 2. No, exit Enter to confirm · Esc to exit
What this means: Claude Code is asking "Can I work with files in your home folder?" This is a normal security check. Since you just installed it yourself, it's safe to proceed.
✅ What to do: The cursor is already on "1. Yes, proceed" (that's the blue text). Simply press Enter on your keyboard.
If this is your FIRST TIME logging in: Your web browser will automatically open and show the Anthropic login page.
What to do:
💡 Already have an account? If you already used Claude on the website (claude.ai), you can use the same email and password here. It's the same account!
🔒 Is this safe? Yes! The Terminal opens your actual web browser (Chrome, Edge, Safari, etc.) and takes you to Anthropic's official website. Your password goes directly to Anthropic, not through the Terminal. It's as safe as logging into any website.
Note: If you've already logged in before on this computer, you might skip this step entirely and go straight to the welcome screen below.
After the browser login (or if you were already logged in), you'll see a welcome screen in the Terminal with an orange border that shows:
What this means: You're logged in and Claude Code is ready! This is like the "home screen" of Claude Code.
✅ You'll also see a prompt at the bottom:
> █This blinking cursor means Claude is listening and ready for your instructions! You can now type messages to Claude, just like chatting.
If Claude shows a message asking about "authentication":
> auth
I need more context to help you. Could you clarify what you'd like to
do with authentication? For example:
- Implement authentication in a project?
- Debug an auth issue?
- Review authentication code?
- Set up OAuth, JWT, or another auth method?
- Something else?Don't panic! This is NOT an error. What happened:
claude auth login earlier, you were talking to your computer's Terminalauth by itself, Claude thinks you want help with authentication coding (like building login systems for websites)✅ What to do: Just ignore this message and move on to testing with a real instruction below. You're already logged in and ready to go!
🎓 Teaching moment: "auth" is short for "authentication" — the technical term for login systems. Programmers use this term when they're building the "Sign In" features you see on websites. Since you're not a programmer (yet!), you won't need this — just type your plain-language requests to Claude instead.
Now let's test that everything is working! You can type anything you want, but here are some examples:
Example 1: Simple test
tell me a joke about spreadsheetsExample 2: Ask about files
what folder am I in right now?Example 3: Get help
what can you help me with?Type any of these (or your own message), press Enter, and Claude will respond! This confirms your installation is complete and working perfectly.
💡 Pro tip: You can chat with Claude in plain English, just like you're texting a colleague. No coding knowledge required for basic questions and tasks!
✅ How to exit Claude Code when you're done: Type exit and press Enter, or simply close the Terminal window.
🔒 Is this safe? Yes! The security prompt is Anthropic's way of making sure you understand Claude Code can work with files in that folder. Since you installed it yourself and it's working in your home directory, it's completely safe to say "Yes, proceed."
Let's make sure everything is set up correctly!
📝 Copy this command:
claude --versionWhat you should see:
claude-code version 1.2.3(The numbers might be different — that's fine! As long as you see "claude-code version" and some numbers, you're good!)
🎉 Congratulations! You're Ready!
If you saw the version number, Claude Code is successfully installed and you're authenticated. You can now use Claude to help with your work!
❌ Didn't work?
If you see "command not found" or similar, go back to Step 3 and make sure you closed and reopened your Terminal. Still stuck? Check the FAQ section below for common troubleshooting tips.
You opened the Terminal (maybe for the first time!), ran commands, and installed professional AI software. That's not "tech stuff" — that's real capability. You're now equipped to use AI tools that most consultants only dream about.
You just completed the installation! Before moving forward, let's make sure you understood the key concepts. This will help you feel confident using Claude Code.
The website (claude.ai) is for chatting. The terminal tool (Claude Code) is for working with your files. They are separate and use different interfaces.
Claude Code runs locally and connects securely to the Claude API for intelligence. Your files remain on your machine; you review all diffs and approve before anything is written.
Store your API key in a password manager. Do not share it, and never save it in a public place like a GitHub repository.
You understand the mental model of how Claude Code works as an API tool and you have a step-by-step plan for installation.
In Module 3, you'll use your newly installed Claude Code to make your first real change to a file and learn the critical "review and approve" workflow.
If you just installed VS Code and aren't familiar with it yet, here's what you need to know:
Don't worry if this sounds confusing now — Module 3 will walk you through every step!