Claude Code en action

Aperçu

Get hands-on with Claude Code, Anthropic's terminal AI agent:

master context, plan mode, custom commands, MCP, and hooks to ship real work you can trust. Turn Your Terminal into an AI-Powered Development Environment Claude Code is Anthropic's agentic coding assistant that lives in your terminal, reading your files, running commands, and editing code on your behalf.

This course shows you how to use it well — not by typing hopeful prompts and accepting whatever comes back, but with deliberate workflows that produce reliable results. You'll work inside a real project the whole way through, and because DataCamp provides a ready-to-use environment, you'll run actual Claude Code sessions from the very first exercise, with no installation or API key required.

Understand How Claude Code Actually Works Before you can guide a tool, you need to know what it is doing. You'll learn what a coding assistant really is, and how a language model — which can only turn text into text — uses tool calls to read files, run commands, and act on the outside world.

You'll see why Claude's strength with tool use is exactly what sets Claude Code apart, and watch it take on real tasks:

profiling and optimizing a popular open-source library, analyzing data in a notebook, restyling a web app through a live browser, and reviewing a pull request for a security issue on GitHub. Build a Reliable Daily Workflow Strong results with Claude Code come down to one core skill:

managing context.

You'll set up Claude on a project with /init and a CLAUDE.md file, point it at the right code using @ file mentions and memory mode, and keep long sessions focused with /context, /compact, and /clear. From there, you'll make safe, reviewable changes with plan mode and extended thinking, commit them with messages Claude writes for you, and turn repetitive prompts into custom slash commands.

Finally, you'll extend Claude Code with MCP servers and wire it into GitHub to review pull requests automatically. Automate and Extend with Hooks and the SDK Claude Code does not have to wait for your next prompt.

You'll learn to write hooks that run automatically before and after Claude uses a tool — blocking reads of sensitive files such as .env, type-checking every edit and feeding errors back so Claude can fix them, and even launching a second Claude instance to catch duplicate code before it lands. Then you'll step outside the terminal entirely with the Claude Code SDK, driving Claude from your own scripts, pipelines, and scheduled jobs.

By the time you finish, you'll know how to set up Claude Code on any project, guide it with just enough context, and extend it with custom commands, MCP servers, hooks, and the SDK — a practical, end-to-end foundation for shipping real work with an AI coding assistant you can trust.

Programme

• Qu'est-ce que Claude Code ?
Découvrez ce qu'est Claude Code, comment il se distingue des autres assistants de codage, et comment son utilisation d'outils lui permet de lire des fichiers, d'exécuter des commandes et d'expliquer le code pour vous.
• Prise en main
Configurez Claude Code sur un projet réel, puis dirigez-le avec CLAUDE.md et des contrôles de contexte, effectuez des modifications sûres avec le mode plan, et étendez-le avec des commandes personnalisées, des serveurs MCP et GitHub.
• Hooks et le SDK
Automatisez Claude Code avec des hooks qui s'exécutent avant et après ses outils, bloquant les lectures de fichiers risquées, vérifiant le type des modifications, et signalant le code dupliqué, puis pilotez Claude depuis vos propres scripts avec le SDK.

---

Enseigné par

Anthropic Academy

---

Matières

Artificial Intelligence