Getting Started: Claude Code

Source: https://dometrain.com/course/getting-started-claude-code/

AI coding assistants are only as effective as the context they operate in, and running them directly in your terminal requires a different approach than a standard web chat. This course teaches you how to effectively integrate Claude Code into your local development environment. You will start by clarifying how Claude Code differs from Claude Chat and Cowork, getting it installed, and navigating the interface. From there, we examine the mechanics of the context window, covering how to avoid context poisoning and how to manage token limits using compaction and the /context command. You will then learn how to configure your projects by writing an effective CLAUDE.md file, structuring modular imports, and understanding how the agent retains memory via the .claude directory.

With your project configured, we move into daily development workflows. You will see how to handle multiple files, execute commands, manage permissions, and resume sessions, culminating in a practical development exercise. Because running autonomous agents can quickly consume tokens, we break down the model lineup, effort levels, and how to use Plan Mode to track and manage your costs. Finally, the course covers integrating Claude Code into your version control system. You will learn how to write commits, create and review pull requests, and manage parallel sessions using Git worktrees, before extending the tool’s capabilities with custom subagents, packaged skills, and MCPs to connect to external systems.