Hello Claude: Your First Conversation

You've installed Claude Code your General Agent. Now it's time to actually use it.

This lesson isn't about learning complex features. It's about having your first real conversation with Claude and realizing it's not intimidating—it's actually natural.

---

The Setup: One Command

Open your terminal and type:

That's it. You're now in Claude Code.

You'll see a prompt. It looks something like:

This is where you talk to Claude. In natural language. No special syntax. No commands you need to memorize.

---

Your First Questions: Orient Yourself

Let's start with the basics. Ask Claude where you are:

Claude will tell you. Your current directory. Simple context.

Now ask what Claude can do:

Claude will explain it naturally. Not a manual. Not technical jargon. Just what's possible.

What you're learning: Claude responds to natural language. You're not "commanding"—you're asking questions and conversing.

---

Real-World Example: Search for News and Save It

Here's something useful. Let's find recent AI news and save it to a file.

Ask Claude:

Watch what happens:

1. Claude reads the web for recent news
2. Claude proposes a file with formatted content
3. You'll see a prompt asking: > Create file ai-news-2026.txt?
4. You press Enter to approve

Done. You now have a file with actual, current information.

What just happened:

• You described what you wanted in plain English
• Claude figured out the steps (search, format, create file)
• Claude showed you what it was about to do
• You approved with one keypress

No clicking. No menu hunting. Just conversation.

---

The Approval Pattern: You're Always in Control

Notice that Claude didn't just do things. It asked first.

This is the permission model. Every time Claude wants to:

• Read a file
• Write or create a file
• Run a command

It shows you exactly what it's about to do and waits for your approval.

Press [Enter] to approve. Press [Esc] to reject.

If you reject, you can ask Claude to try differently. That's steering.

Why this matters: You're never surprised. Claude can't accidentally delete your project or run something dangerous without your knowledge.

---

Ready to Practice?

Head to Lesson 06: Practical Problem-Solving Exercises for 27 hands-on exercises with complete walkthroughs. You'll organize messy files, analyze data, create documents, and build problem-solving skills — all with one-click exercise downloads and step-by-step guidance.

---

Next: Extending Claude (Preview)

Once you master basic conversations, you'll start asking more ambitious questions:

• "How do I make Claude repeat this task every morning?"
• "How do I connect Claude to my Jira board so it can check for new tickets?"
• "How do I enforce my team's naming conventions automatically?"
• "How do I give Claude access to my company's database?"

This is where Claude's extension tools come in. You don't need to understand them yet—just know they exist.

The Five Extension Tools at a Glance

Why This Matters Now

As you continue through the chapter, you'll recognize patterns:

• "I need Claude to always format code like this" → You'll think: That's a CLAUDE.md use case
• "I want Claude to research and write a report without interrupting me" → That's a Subagent
• "I need Claude to check GitHub for vulnerabilities before I commit" → That's a Hook

You're building a mental model. The tools are the same; the architecture is what separates amateur automation from professional systems.

---

Try With AI

Explore naturally:

"Ask Claude: 'What's in this directory?' Then ask: 'Tell me about my project structure.' Watch how Claude learns about your environment through conversation."

Create something real:

"Ask Claude to search for recent breakthroughs in your field of interest and save them to a file. Use natural language—describe what you want, not how to do it."

Test the approval workflow:

"Ask Claude to create a file, and when it asks for approval, press Esc to reject. Then ask it to do something different. Notice how Claude adapts without frustration."

Next Up: Now that you're comfortable conversing with Claude, let's teach Claude about YOUR preferences.