Open your terminal and navigate to an empty folder. This is where we'll build everything.
claude
What you learned: Claude Code runs in your terminal. Just type claude to start a conversation.
Slash commands are special instructions that control Claude Code itself (not prompts for Claude).
/help
What you learned: Slash commands start with / and control the CLI. Regular text is your prompt to Claude.
These commands help you understand your current session state.
/login # Authenticate with Anthropic
/status # Check connection status
/model # See/change current model
/usage # View token usage this session
/config # View/edit configuration
/stats # Detailed usage statistics
💡 Run /status anytime to check if you're connected and which model you're using.
Press Shift+Tab to cycle through three modes that change how Claude behaves.
Default Mode
Claude asks permission before each file change. Safest for learning.
Auto-Accept Mode
Claude makes changes without asking. Faster, but review carefully!
Plan Mode
Claude creates a plan first, then executes after your approval.
What you learned: Shift+Tab switches modes. Start with Default, use Auto-Accept when confident.
Understanding how Claude remembers your conversation.
/context # See what's in Claude's context window
/compact # Summarize conversation to free up space
/init # Create CLAUDE.md for project memory
📖 Context is everything Claude can "see" right now. If it fills up, run /compact to continue.
Save your work and pick up where you left off.
/resume # Resume a previous session
/exit # End the session (or Ctrl+C)
What you learned: Your sessions can be saved and resumed. Use /resume to continue from where you left off.
Let's set up CLAUDE.md so Claude remembers your project preferences. Use this when you are opening any existing project.
/init
What you learned: /init creates a CLAUDE.md file. Claude reads this every session to understand your project.
Now let's actually talk to Claude. Ask it something simple.
What files are in this directory?
📖 Claude Code can read files, write code, run commands, and help you build entire projects through conversation.
Let's make something. Ask Claude to create a simple file.
Create a file called hello.txt with the text "Hello from Claude Code!"
What you learned: Claude doesn't just suggest code — it actually creates files for you. Check your folder!
Use @ to reference files directly in your prompts. Claude can read PDFs, images, and code files.
@linkedin.pdf summarize my professional profile
@screenshot.png what's in this image?
@src/app.js explain this code
📖 The @ symbol tells Claude exactly which file to look at. Works with PDFs, images (PNG, JPG), code files, and more!
What you learned: Use @filename to reference specific files. Claude can read and understand PDFs, images, and code.