Claude Code Basics · the big idea
It's not ChatGPT in a terminal
A chatbot hands you words and code to copy somewhere yourself. Claude Code works inside your actual project — it opens your files, makes the changes, and runs the commands for you.
💬A chatbot tells you — "here's the code, now go paste it." You're still the one doing the work.
🤖Claude Code does it — it edits the real files in your real project and shows you the result.
That's the difference between a helper that talks and a builder that acts.
The one idea the whole course hangs on
It works in a loop: read → act → verify
This loop is what makes it an agent, not just a chat. It keeps going until the job is actually done.
1ReadsLooks at your project — the files, the structure — to understand what's really there before touching anything.
2ActsMakes the change: writes or edits files, runs the commands needed to build it.
3VerifiesChecks its own work — runs it, reads the output — and loops back to fix what's off.
…and you're in every loop
You stay in the loop
You set the goal and approve the moves. Claude does the heavy lifting and shows its work — you're the one steering.
🎯You point — say what you want in plain language. No jargon required.
✅You approve — it asks before changing things; nothing happens behind your back.
🪟Your window, your call — a friendly panel in your editor or the terminal. Same builder either way.
Read → Act → Verify. Remember this.
Everything else in this course is just getting better at steering this loop.
← Back to course
