Course
Build Your First AI Coding Agent
From zero configuration to a running code review agent in six hands-on lessons.
About this course
<br />
You will learn what agents are and when to use them, set up CLAUDE.md for your project, create custom skills as slash commands, connect MCP servers to external tools, and wire all three layers into an automated review workflow.<br />
<br />
Designed for developers who have heard of Claude Code but have not configured an agent before. No prior AI coding tool experience required.
What you'll learn
Course lessons
Prerequisites
Frequently Asked Questions
Each lesson is self-contained and can be read independently. However, lessons 3 through 6 build on each other. If you already understand what agents are and how Claude Code's system works, you can skip to lesson 3 and start configuring.
The course focuses on Claude Code because it has the most structured agent system (CLAUDE.md, skills, MCP, hooks). Many of the concepts transfer to other tools. Skills use a portable format that works across Cursor, Codex CLI, and Gemini CLI.
The course content is free. Using Claude Code requires a Pro subscription at $20/month minimum. The exercises use minimal tokens throughout. A single session budget covers all six lessons comfortably.
You will have a working code review agent and the knowledge to build other agents using the same three-layer pattern. Common next projects include deployment agents, documentation updaters, and bug triage systems that pull from Sentry or Linear.