An intro and six parts covering what Claude Code actually is, how to set it up and configure it well, and how to work with it effectively in production. Required pre-reading for Track 3 — Developer workshop.
Intro · Claude Code Fundamentals Beginner
Claude Code is a local AI agent you run in your terminal. It is not a web chat interface and not an IDE plugin. Before you can use it, you need three things — a Node.js environment, an Anthropic account, and one install command. This guide covers all of them.
Part 1 of 7 · Claude Code Fundamentals Beginner
Most developers pick up Claude Code expecting it to behave like a faster Copilot. It does not. It is an agentic loop that reads, acts, observes, and repeats. That distinction changes everything about how you use it.
Part 2 of 7 · Claude Code Fundamentals Beginner
VS Code is the recommended starting environment for Claude Code. The extension makes every step of the agentic loop visible — plans you can review, diffs you approve, checkpoints you can rewind to. Here is what to set up and how to work with it.
Part 3 of 7 · Claude Code Fundamentals Intermediate
The engineers who get the most from Claude Code invest in configuration, not prompting. CLAUDE.md encodes project knowledge. Skills encode process. Hooks enforce policy. Together they make the agentic loop consistent without re-explaining context every session.
Part 4 of 7 · Claude Code Fundamentals Intermediate
Claude Code can spawn multiple agents running in parallel, connect to external systems via MCP, and integrate into team workflows at scale. This is the same architecture that powers the Code Review feature — and it is available in every session.
Part 5 of 7 · Claude Code Fundamentals Advanced
Permission modes are not a convenience setting. They define how much the AI acts before you review — which means they define who owns the consequences. Choosing the right mode for the right context is one of the most practical decisions a developer makes with Claude Code.
Part 6 of 7 · Claude Code Fundamentals Advanced
Context is not free. It accumulates with every message, carries your CLAUDE.md on every request, and degrades when it grows too large. Understanding how tokens flow through a session — and how to design sessions that stay efficient — is a production skill, not an advanced niche.
Capstone · Claude Code Fundamentals
How the Code Review feature deploys the fleet subagent architecture in production — and what you still own as the engineer.
Track 3 pre-task: Read Parts 1–3 before the developer workshop. Parts 4 and the capstone article are covered during the session.
Book Track 3 workshop