Chapter 1: Ground Zero
Your First Conversation
Your First Conversation
Time to talk to Claude Code! Let's start with something simple to see how it works.
Start Claude Code
Open your terminal and navigate to any folder:
Claude Code starts an interactive session. You'll see a prompt where you can type natural language instructions.
Try these prompts
Here are some things you can ask right away:
- Ask a question: "What files are in this directory?"
- Create a file: "Create a hello.js file that prints Hello World"
- Run a command: "Run the hello.js file with node"
- Modify code: "Change it to accept a name as a command line argument"
Notice how specific those prompts were. "Create a hello.js file that prints Hello World" is much better than "make me a file." This is the golden rule of prompting: if you showed your instruction to a colleague and they'd be confused, Claude will be confused too. Be clear, be direct, say exactly what you want.
Notice how Claude Code doesn't just show you code — it actually creates and modifies files on your system. It asks for permission before running commands.
Understanding the flow
When you give Claude Code an instruction, it:
- Thinks about what needs to be done
- Reads relevant files if needed
- Plans its approach
- Executes — writing files, running commands
- Reports what it did
Claude Code handles ambiguity, but specific prompts get better results on the first try. "Add a GET endpoint at /api/users that returns a JSON array" will land closer to what you want than "Build me a REST API." When in doubt, be specific.
Key shortcuts
- Ctrl+C — cancel the current operation
/help— see available commands/clear— clear conversation history- Up arrow — cycle through previous prompts
Great work! You've had your first conversation with Claude Code. Next, we'll scaffold a real project.