What is Claude Code?

VIDEO PROCESSING...

Module outcome

By the end of this module, you will be able to explain what Claude Code is, describe what kinds of tasks it can tackle, and choose the right surface for your workflow.

Claude Code is an agentic coding tool built by Anthropic. It reads your codebase, edits files, runs commands, and integrates directly into your development workflow — all from a single conversation.

Why it matters: Traditional AI coding tools generate snippets you paste. Claude Code is fundamentally different — it operates inside your project, reads your actual files, understands your architecture and takes action autonomously. It's the difference between a search engine and a coworker.

Key insight

Claude Code is not a chatbot. It's an agent — it takes actions in your environment (reads files, edits code, runs commands) instead of just answering questions.

How the agentic loop works:

When you give Claude Code a task, it doesn't just respond once. It runs a loop:

1. Read — explores your codebase, reads relevant files, understands the context
2. Plan — decides what steps are needed to complete the task
3. Act — edits files, runs commands, executes tests
4. Verify — checks the results, fixes errors, iterates

This loop continues until the task is done or it needs your input. You stay in control — you can approve, correct or cancel at any step.

Claude Code vs other tools:

Tool	What it does	What it can't do
GitHub Copilot	Autocomplete inside your editor	Run commands, read your full codebase, take multi-step actions
ChatGPT / Claude.ai	Answer questions, generate code snippets	Access your files, run your tests, create real PRs
Cursor	AI-enhanced editor with inline generation	Autonomous multi-step execution, CI/CD integration, pipe automation
Claude Code	All of the above + acts autonomously in your environment	—

The key distinction: Claude Code has access to your environment. It doesn't just see the code you paste — it explores your actual repo.

What "reads your codebase" actually means:

When you start Claude Code in a project, it can:

+Read any file you reference or that's relevant to the task+Understand how files relate to each other (imports, exports, dependencies)+Check your git history to understand recent changes+Run commands to discover the project structure+Look at test results and error messages to diagnose problems

This is context awareness at the project level — not just the file you have open.

bash

1# Claude Code covers your full dev workflow:
2claude "explain what this project does"         # Understand
3claude "fix the login bug and run tests"        # Fix
4claude "write tests for the auth module"        # Test
5claude "create a PR with these changes"         # Ship
6
7# It also handles compound tasks in one prompt:
8claude "find all TODO comments, create GitHub issues for each, then delete the comments"

A concrete example:

You just merged a dependency update and the CI is red. In the old world: you open the CI logs, read the error, search Stack Overflow, paste snippets, test locally, iterate.

With Claude Code:

```
> "CI is failing after updating lodash. Find the error, fix it, and run tests."
```

Claude reads the failing test output, traces the error to a breaking API change, edits the affected files, runs npm test locally to verify, then reports back with a summary of what it changed and why. Total time: under 2 minutes.

Your turn

Open a terminal and run: claude "what is 2+2?" — just to confirm Claude Code responds. If you don't have it installed yet, that's what Module 2 is for. No project needed for this check.