This course covers all the things that I have learned about using AI-powered tools like Cursor and Claude Code to write software. Since this kind of stuff can go out of date quickly, I try to focus on the battle-tested approaches that have worked for decades: thinking things through before you jump into the code, maintaining good Git discipline so you can revert back to a known good state whenever things go off the rails, testing, linting code quality, and more.
We also spend a decent amount of time accidentally learning how large language models work and how to effectively manage context.
And, of course, we’ll actually cover the tools themselves. You’ll learn to use advanced AI coding assistants like Claude Code and Cursor, understand the Model Context Protocol (MCP) for extending AI capabilities, and master patterns that ensure consistent, high-quality results.
These are my notes for my course for Frontend Masters and are designed to be supplementary to that course. If they’re helpful to you outside of the course, then that’s an added bonus. There is also plenty of content that I won’t get to in the course. So, these notes serve as a way to assuage my guilt about the very practical trade-offs that come from the fact that you can only fit so much stuff into a single day.
Resources and References
- Basic React, TypeScript, and Tailwind Template
- Model Context Protocol Overview
- Repomix - Tool for preparing repository content for AI analysis
- Extended Thinking Mode (Anthropic) - Documentation on Claude’s thinking capabilities
- cursor.directory - Community examples for Cursor rules
- awesome-cursorrules - Collection of Cursor rules examples
- Crystal: A tool for managing Git worktrees for Claude Code.
- Claude Code Sub-Agents Collection
- Awesome Claude Prompts
- Awesome Claude Code