Module Introduction: Key Features & Efficient Usage
You've set up Claude Code. You know how sessions, context, and permissions work. Now comes the part that actually changes how you build software.
What this chapter covers
This chapter is about the features that separate Claude Code users from Claude Code engineers. We'll go deep on:
Context Engineering How to write prompts and structure inputs so Claude's output is consistently high quality — not just occasionally good.
CLAUDE.md Files How to write a system prompt that makes every session feel like working with a colleague who already knows your project.
Plan Mode How to make Claude think before it acts — especially important for complex, multi-file changes.
MCP Servers How to connect Claude Code to external tools: databases, APIs, file systems, and custom services.
Subagents How to decompose large tasks into parallel workstreams, each with their own fresh context.
Agent Skills Reusable, callable "skills" that Claude can invoke — like functions for your AI workflow.
Hooks and Plugins Automated quality control at every event, plus extending Claude Code's built-in toolset.
Loops Running Claude Code autonomously in a self-correcting feedback cycle.
Learning approach
Every concept is explained with a real project context — not toy examples. By the end of this chapter, you'll have a complete mental model of how all these features fit together into a professional agentic engineering workflow.