Claude Code Tutorial — Complete Guide

A complete, hands-on Claude Code tutorial series. Learn Claude Code from first install to advanced agentic workflows — CLAUDE.md, hooks, skills, MCP servers, subagents, plugins, and production AI systems.

Claude Code Tutorial — Complete Guide

Claude Code is Anthropic’s agentic CLI — it reads your codebase, writes files, runs commands, and coordinates multi-agent workflows, all from your terminal. This series takes you from your first install to building production AI systems. Every article is practical, covers real use cases, and builds on the one before it.

21 articles. Every layer of the stack. DevOps-focused throughout.

Part 1: Foundations

Before writing a single config file, understand the model you are working with and how the CLI is structured.

  1. Claude Models in 2026: Opus, Sonnet, and Haiku Compared
  2. Mastering Claude Code CLI: The Complete Guide for DevOps Engineers

Part 2: Configuring Claude’s Behaviour

Claude Code is only as good as its configuration. These articles cover the two most important levers — the project-level instruction file and the full automation layer — plus the checkpoint system that protects your work.

  1. Writing a CLAUDE.md That Actually Works
  2. Claude Code Hooks, Commands, Skills, and Subagents: The Complete Guide
  3. Claude Code Skills: Build Reusable AI Workflows for Your Team
  4. Claude Code Sub-Agents: Specialist Assistants and Delegation
  5. Claude Code Session Management: Parallel Work and Persistent Context
  6. Claude Code Checkpoints: Your AI Safety Net for Code Changes

Part 3: Extending Claude Code

Connect Claude to external systems, package your configuration for reuse, and put it in a continuous loop.

  1. MCP Servers Worth Installing: For Developers, Testers, and DevOps Teams
  2. MCP Protocol Deep Dive: How Claude Code Connects to External Tools
  3. Claude Code Plugins: The Complete Guide to Building and Sharing Extensions
  4. The Claude Code /loop Command: In-Session Automation Explained

Part 4: Performance and Cost

Left unchecked, Claude Code sessions get expensive. These two articles cover every technique for keeping costs low without sacrificing capability.

  1. Stop Burning Tokens: A Practical Guide to Claude Code Cost Optimization
  2. Claude Prompt Caching: Cut Your API Costs by 90%

Part 5: Advanced Capabilities

Push Claude beyond code generation — voice input, extended reasoning, UI automation, and security auditing.

  1. Voice Dictation in Claude Code: Speak Instead of Type
  2. Claude Extended and Adaptive Thinking: Making Claude Reason Before It Answers
  3. Claude Computer Use for DevOps: When to Use It and When to Use an API
  4. Claude Code as a Security Scanner: Beyond Pattern Matching

Part 6: Building AI Systems

Graduate from using Claude to building with it — autonomous agents, hosted infrastructure, and no-framework approaches.

  1. Claude Managed Agents: Deploy AI Agents Without Managing Infrastructure
  2. Building a Personal AI Assistant with Claude Agent SDK and Bun
  3. You Don’t Need a Framework to Build an AI Assistant

What You Will Learn

  • How the Claude model family works — when to use Opus, Sonnet, and Haiku, and how to route by task
  • Every CLI flag, slash command, and keyboard shortcut in Claude Code
  • CLAUDE.md as behavioral programming — not documentation, but runtime instructions
  • The full automation hierarchy: hooks → commands → skills → subagents and when to reach for each
  • How to connect Claude to databases, Kubernetes clusters, GitHub, and any API via MCP
  • How to package your configuration as a shareable plugin
  • Five habits that cut token usage by up to 50%
  • Prompt caching and the Batch API — together they reduce API costs by up to 95%
  • Extended thinking and adaptive thinking — what they are and when to turn them on
  • Computer use: when GUI automation beats a structured API integration
  • How to run Claude as a security auditor that finds logic flaws, not just lint warnings
  • Managed Agents: the hosted REST API that eliminates agent infrastructure work
  • Two approaches to building autonomous AI assistants — SDK-powered and shell-script-powered

Prerequisites

  • Basic terminal and shell comfort (cd, git, environment variables)
  • Any programming language experience — examples use TypeScript, Python, and Bash
  • An Anthropic API key (or Claude.ai subscription for the CLI)

No prior Claude or AI experience required. Parts 5 and 6 assume you have worked through Parts 1–4.