Claude Code

Claude Code is Anthropic's AI agent that works on your computer: it reads your project's files, runs commands and makes changes over several steps, asking your permission for what matters. You use it from the terminal, the desktop app or editors like VS Code.

Select any words here to ask your Maestro about them.

You asked: "The total in sales.csv doesn't match the report. Find out why." It took these six steps on its own. Before which ones do you think it asked you?

A sample session with the default settings: it reads without asking and asks before it edits files or runs commands. You decide what else it may do without asking. Ring: it asked you there.

What it's for: handing off multi-step work on your own files: cleaning a CSV, writing and running a Python script, building a report, tracking down why a total doesn't tie out. It was built for programming, but data people use it more and more for analysis, because it works where your data lives.

It's an AI agent: an LLM from Anthropic (Claude) inside a harness that gives it tools to read and write files and run commands in your terminal. It asks before it edits files or runs commands, except for what you allow it to do without asking; it reads its agent instructions file (CLAUDE.md) when it starts, and can load skills and connect to other tools over MCP.

You use it from the terminal, from the Claude desktop app or inside editors like VS Code. It needs a paid Claude account or an API key. To work on your files without a terminal, Claude Cowork does something similar from the app.

In the course Agentic AI for data, you use it (or Codex CLI, or Antigravity CLI) to analyze data without giving up your judgment.

How it's used

A first session with a CSV

sales-2026/

$ cd sales-2026
$ claudeOpens Claude Code in this folder. What it can see is what's in here.
> Read sales.csv and tell me what a row is. Don't change anything.What you type is a prompt, just like in a chat.
● Read(sales.csv)It uses a tool: it actually opens your file instead of guessing what's in it.
● Bash(wc -l sales.csv)It runs a command in your terminal to count lines. It asks before running it, unless you've already allowed it.
marked: Each row is one item on a ticket: 4,811 rows across 1,284 tickets.

The difference from a chat: here the model opens your files and runs commands on your computer. That's why you can check what it did, step by step.

When to look closer

Chat or agent?

What Claude Code does

Answers questions in plain language
Opens the files in your folder on its own
Runs commands and scripts on your computer
Reads a file of your rules (CLAUDE.md) when it starts

Which of these four can't a chat like ChatGPT in the browser do?

Ask your Maestro about Claude Code

It answers from the free course page closest to this word, the way it answers while you study.

Example

You: What does "Claude Code" mean, in plain words?

Your Maestro: I'll explain it with a case of yours: it sticks better than a definition. What job on your files would you hand an agent first?

Asking about: Answers from the page "Why does my CSV use semicolons?"

Only this browser keeps what you ask here. How it works

Learn more

Claude Code docs

AI agent
A program that pairs a language model with tools, such as terminal commands and files, to carry out a task in several steps: it reads its context, acts, looks at the result and repeats.
Agent harness
The program that wraps a language model and makes it an agent: it gives the model tools, asks for your permission, keeps the session and decides what goes into its context.
Agent instructions file
A text file in your project that the agent's harness loads at the start of every session, so the agent knows from the start what it always needs to know: rules, facts about the data and how to check its work.
Agent skill
A skill is a folder with a SKILL.md file that teaches an AI agent a procedure: when to use it (its description) and how to do it. The agent reads only the description up front and loads the rest when a task calls for it.

The whole glossary