How to Use Claude Code: Complete Beginner Guide (2026)

Claude Code is Anthropic's official CLI tool that lets you use Claude directly in your terminal to write, edit, and debug code. Unlike ChatGPT or Claude.ai (which run in a browser), Claude Code works inside your project directory -- it can read your files, edit them, run commands, and build features without you ever leaving the terminal.

This guide covers everything you need to get started: installation, setup, your first task, and advanced workflows. No prior experience with AI coding tools required.

In this guide:

What Claude Code does
Installation and setup (5 minutes)
Your first Claude Code task
Common commands and workflows
Tips for better results
Connecting to a multi-agent squad

What Claude Code Does

Claude Code is a terminal-based AI coding assistant. You type natural language instructions, and Claude Code:

Reads your project files to understand your codebase
Writes and edits code across multiple files
Runs terminal commands (tests, builds, git operations)
Debugs errors by reading error messages and fixing code
Explains code in plain English

Unlike browser-based AI tools, Claude Code has direct access to your file system. It can see your entire project, make edits, and verify changes by running commands.

What Claude Code is NOT: It's not a code editor with an AI chat panel (that's Cursor). It's not a browser chatbot (that's Claude.ai). It's a terminal-native tool for developers who prefer the command line.

For a detailed comparison with other tools, see our Claude Code vs Cursor comparison or our three-way comparison of Cursor vs Windsurf vs Claude Code.

Installation and Setup

Prerequisites

Node.js 18+ (check with node --version)
An Anthropic API key (get one here)
A terminal (Terminal on macOS, PowerShell or Git Bash on Windows)

Step 1: Install Claude Code

npm install -g @anthropic-ai/claude-code

This installs the claude command globally. Verify it works:

claude --version

Step 2: Set Your API Key

Claude Code needs your Anthropic API key to communicate with Claude. Set it as an environment variable:

macOS/Linux (add to ~/.bashrc or ~/.zshrc):

export ANTHROPIC_API_KEY="sk-ant-your-key-here"

Windows PowerShell:

$env:ANTHROPIC_API_KEY = "sk-ant-your-key-here"

Or set it inline when running:

ANTHROPIC_API_KEY="sk-ant-your-key-here" claude

Get your API key from the Anthropic Console. New accounts get $5 in free credits. For typical Claude Code usage, $5 lasts 1-2 weeks of daily use.

Step 3: Navigate to Your Project

Claude Code works best when you run it inside a project directory. It reads your files to understand context.

cd your-project-folder
claude

You'll see the Claude Code prompt:

claude>

You're ready to go. Total setup time: about 3 minutes.

Your First Claude Code Task

Example 1: Ask Claude Code to Explain Your Project

claude> Explain what this project does and how it's structured

Claude Code will read your files (package.json, README, main source files) and give you a summary. This is a good first task because it shows you how Claude Code understands your codebase.

Example 2: Ask Claude Code to Write a Feature

claude> Add a /health endpoint to the Express server that returns { status: "ok", uptime: <seconds> }

Claude Code will:

Find your Express server file
Read the existing code structure
Add the health check endpoint
Show you the changes it made

Example 3: Ask Claude Code to Fix a Bug

claude> The login form doesn't show error messages when the password is wrong. Fix it.

Claude Code will find the login component, read the form handling logic, identify where error handling is missing, and add it.

Common Commands and Workflows

File Operations

Claude Code can work with any file in your project:

claude> Read src/utils/auth.ts and add input validation to the login function
claude> Create a new file src/middleware/rateLimit.ts with Express rate limiting
claude> Find all files that import the old API client and update them to use the new one

Running Commands

Claude Code can execute terminal commands:

claude> Run the test suite and fix any failing tests
claude> Run npm run build and fix any TypeScript errors
claude> Check the git status, stage all changes, and commit with a descriptive message

Multi-File Refactoring

One of Claude Code's strengths is making changes across multiple files:

claude> Rename the getUserById function to findUserById across the entire codebase. Update all imports, tests, and references.

Claude Code will search for every occurrence, update files, and verify nothing breaks.

Git Workflows

claude> Look at the git log for the last 5 commits and write a release summary
claude> Create a new branch called feature/user-preferences, implement a user preferences page, and open a PR

Debugging

claude> The app crashes on startup with ECONNREFUSED. Read the error, find the cause, and fix it.
claude> These 3 tests are failing. Read the test output, find the bug, and fix it without changing the test expectations.

Tips for Better Results

1. Be Specific About What You Want

Vague: "Make the code better" Specific: "Add error handling to the fetchUserData function. If the API call fails, show a user-friendly error message and retry up to 3 times with exponential backoff."

2. Provide Context

claude> This is a Next.js 15 app using the App Router. Add a /dashboard route that shows the logged-in user's recent activity. Use the existing useAuth hook from src/hooks/useAuth.ts.

3. Ask Claude Code to Plan Before Acting

claude> I need to add Stripe payments to this app. First, read the codebase and tell me your plan. Don't make any changes yet -- just describe what you'd do.

This lets you review the plan before Claude Code starts editing files.

4. Use CLAUDE.md for Project Context

Create a CLAUDE.md file in your project root with project-specific instructions:

# Project: My SaaS App

## Tech Stack
- Next.js 15 (App Router)
- PostgreSQL with Prisma ORM
- Tailwind CSS
- Deployed on Vercel

## Conventions
- Use TypeScript strict mode
- All API routes go in src/app/api/
- Database queries use Prisma client from src/lib/db.ts
- Never modify migration files directly

## Testing
- Run tests with: npm test
- Test files go next to the source file: component.tsx → component.test.tsx

Claude Code reads CLAUDE.md automatically when it starts. This gives it context about your project conventions, tech stack, and preferences.

5. Cost Control

Claude Code uses your Anthropic API key. Typical costs:

Activity	Approximate Cost
Simple question about a file	$0.01-0.03
Add a feature to one file	$0.05-0.15
Multi-file refactoring	$0.10-0.30
Full bug investigation + fix	$0.05-0.20
Complex feature (5+ files)	$0.20-0.50

Most developers spend $2-5/day with regular use. Use our AI cost calculator to estimate your monthly costs. For a full pricing comparison across coding tools, see our Copilot vs Cursor vs Windsurf comparison.

Connecting to a Multi-Agent Squad

Claude Code works great on its own. But if you want to coordinate Claude Code with other AI agents -- for example, having a Research agent gather requirements, a Coder agent (Claude Code) implement features, and a Reviewer agent check the output -- you can connect it to Ivern Squads.

How It Works

Sign up at ivern.ai/signup (free, no credit card)
Add your Anthropic API key in Ivern settings (BYOK -- zero markup)
Create a squad with a Coder agent connected to your Claude Code instance
Assign tasks from the Ivern dashboard -- they route to Claude Code in your terminal

Your Claude Code agent works alongside other agents (OpenAI, Gemini, Cursor agents) in the same squad. Each agent handles its specialty while a task board tracks progress.

For the full setup walkthrough, see our Claude Code + Ivern integration tutorial and our guide to connecting Claude Code to a task board.

Why Use Multi-Agent Instead of Claude Code Alone?

Claude Code is excellent at writing and editing code. But real development work involves more than coding:

Research: Understanding requirements, finding APIs, reading documentation
Implementation: Writing the actual code (this is where Claude Code excels)
Review: Checking for bugs, security issues, performance problems
Testing: Writing and running tests to verify correctness

With Ivern Squads, each phase gets a dedicated agent. Your Claude Code agent handles implementation while other agents handle research, review, and testing. See our autonomous AI agent examples for real workflow breakdowns.

Frequently Asked Questions

How do I install Claude Code?

Run npm install -g @anthropic-ai/claude-code in your terminal. You need Node.js 18+ and an Anthropic API key from console.anthropic.com. Set the key as an environment variable: export ANTHROPIC_API_KEY="your-key". Then run claude in any project directory. The full setup takes about 3 minutes.

Is Claude Code free?

Claude Code itself is free to install. You pay for API usage through your Anthropic account. New accounts get $5 in free credits. Typical usage costs $2-5/day for a developer working full-time. This is cheaper than Cursor Pro ($20/month) for developers who already have an Anthropic API key. See our AI coding tool cost comparison for detailed pricing.

Can Claude Code edit multiple files at once?

Yes. Claude Code can read, create, and edit multiple files in a single task. For example, you can ask it to "add a user preferences feature" and it will create the route, update the database schema, add the UI component, and update the navigation -- all in one session.

What is the difference between Claude Code and Cursor?

Claude Code is a terminal tool -- you type commands in your CLI. Cursor is a code editor (fork of VS Code) with an AI chat panel. Claude Code is better for developers who prefer the terminal. Cursor is better for developers who want an IDE-like experience with AI. See our detailed Claude Code vs Cursor comparison for the full breakdown.

Can I use Claude Code with other AI tools?

Yes. Through Ivern Squads, you can connect Claude Code to OpenAI agents, Gemini CLI, Cursor agents, and OpenCode agents in coordinated squads. Each agent handles its specialty while a shared task board manages workflow. See our multi-agent coding setup guide for instructions.

How do I get better results from Claude Code?

Three tips: (1) Be specific about what you want -- include file names, function names, and expected behavior. (2) Create a CLAUDE.md file with project conventions and tech stack details. (3) Ask Claude Code to plan before acting -- say "describe your plan first, don't make changes yet." This lets you review before edits happen.

Does Claude Code work with Python projects?

Yes. Claude Code works with any programming language and project type -- Python, JavaScript, TypeScript, Rust, Go, Ruby, Java, and more. It reads your files, understands the project structure, and writes code in whatever language your project uses.

What models does Claude Code use?

Claude Code uses Claude 3.5 Sonnet by default (Anthropic's recommended model for coding tasks). The model is configured automatically -- you don't need to select it. API costs reflect Claude 3.5 Sonnet pricing: $3/million input tokens, $15/million output tokens. A typical coding task uses 2,000-5,000 input tokens and 1,000-3,000 output tokens.

Next Steps

Install Claude Code: npm install -g @anthropic-ai/claude-code
Get your API key: console.anthropic.com
Try it on a real project: cd your-project && claude
Optional: Connect to Ivern Squads for multi-agent coordination

Start building with AI agent squads -- free signup