Claude Code in Action: The Loop, for Real

In short

Here you just watch: nothing to install, nothing to do. Hands on the keyboard come a couple of chapters from now.
What you see scrolling by isn't magic: it's the loop from the previous lesson, move by move.
You'll also see an error, and that's a good sign: it's part of the loop.
The goal isn't to understand the result — it's to recognize the mechanism.

What we're doing

We open Claude Code in an empty folder and ask it for exactly one thing: a small dashboard that shows how a channel's numbers are doing. A one-line request, the kind you'd give a new hire on their first day.

Then we don't touch anything else. We just watch the rest.

The final result doesn't matter, and it doesn't matter whether you understand what it writes: what matters is that, as it scrolls by, you recognize the pieces we talked about in the previous lesson. If you recognize them, you've understood Claude Code — the rest of the course is just details on top of this.

What to watch for, as it runs

There are five moments. Each one is a move in the loop:

It looks around. Before writing anything, it goes and checks what's in the folder. It doesn't know, so it asks: that's the first tool request.
It says what it's about to do. Every so often a line of reasoning shows up. That's the "think" move: it's not decoration, it's the moment it decides the next step.
It writes files. It doesn't hand you text to copy: it actually creates the files in your folder. This is where the difference with chat shows.
It runs a command and something goes wrong. A red error, a result that doesn't add up. That's not a failure: the result comes back and becomes the information it uses to pick the next move.
It stops and talks to you. At some point it stops asking for tools and writes a normal message. That's the signal that, as far as it's concerned, the work is done.

What shouldn't worry you

Watching a session for the first time, three things tend to spook people — and all three are normal:

The speed. A lot of text scrolls by. You don't have to read all of it: what matters shows up in the moves, not in the lines.
The errors. It makes them. It notices them and fixes them on its own, and that's exactly why the loop exists.
The code. If you can't read it, that's fine: you're not learning to write code, you're learning to direct someone who writes it.

What's next

You've seen the loop run. But there's one thing you can't see from outside, and it's the thing that explains everything else: at every move, the model has forgotten every move before it. In the next lesson we figure out how it still works anyway.

Did you get it?

Two minutes, no grade. Trying to answer — even getting it wrong — cements the concepts far more than rereading. Not in the mood? Skip it: the lesson still counts as complete.

In the video, Claude Code gets a step wrong and tries again. What does that mean?

Do you need to have Claude Code installed to follow this lesson?

1
1. 1
2. 2
3. 3
2
1. 4
2. 5
3. 6
4. 7
3
1. 8
2. 9
3. 10
4. 11
5. 12
4
1. 13
2. 14
3. 15
4. 16
5. 17
6. 18
5
1. 19
2. 20
3. 21
4. 22
5. 23
6. 24
6
1. 25
2. 26
3. 27
4. 28
5. 29
6. 30
7
1. 31
2. 32
3. 33
4. 34
8
1. 35
2. 36
3. 37
4. 38
9
1. 39
2. 40
3. 41
10
1. 42
2. 43
3. 44
11
1. 45
2. 46