🤖The Modern AI Developer
Meet AI & Agents
No background needed: what an Agent is, how it thinks, and a hands-on first run where it actually calls a tool for you.
Context Engineering: Feed the AI Exactly the Right Info
How well AI answers usually depends less on how 'smart' it is and more on what you let it see. This level teaches context engineering: putting just the right information into the AI's 'working memory'—bring the docs it needs, feed back the errors it should see, set the role clearly, write the spec plainly—so the same model does far better work.
Letting AI Actually Do Things: AI IDEs & Tooling
In the first two levels the Agent could only 'think.' This level lets it actually act: working in an AI IDE (Cursor / Claude Code) on your project, learning the sync and async ways to collaborate, how to let go, and how to keep long context from derailing it. Then on to code-gen patterns—designing good tools for it, using CLAUDE.md to make it remember the project's rules, turning personal tricks into team conventions, and finally plugging in MCP to add a whole row of external capabilities at once.
Configuring Your AI Terminal
A modern AI terminal (command line, or the AI assistant in your IDE) isn't a set-and-forget thing—how much it's allowed to do, and how cautious it is, is up to you. This level walks through the product-design principles behind these tools, distinguishes the cautious and aggressive configuration styles, and teaches you to use 'cost of a mistake + reversibility' as a ruler for deciding when to let it run fully automatic and when to require a human checkpoint.
Don't Let AI Cause Trouble: Testing & Security
Your Agent reads email, clicks links, calls tools, runs code—the more it can do, the more ways it can cause trouble. This level explains it all in the most everyday terms: how someone can trick your Agent with a single sentence (prompt injection), why secrets must never go into the prompt, why you can't fully trust the AI when it says 'security scan passed,' and how to back up the AI's output with automated tests plus human review. By the end you'll have a whole set of 'don't let AI cause trouble' defensive instincts.
Modern Code Review
AI wrote you a big pile of code—but do you really know how to review it? This level takes you from zero through the three layers of code review, the key differences between reviewing AI code and human code, and how to split the work sensibly between you and the AI—which mechanical checks to hand off to the AI, and which context-dependent judgments to keep for yourself.
Build a Real App with AI
Learn to turn a fuzzy idea into a real web page anyone can open and use in a browser, with AI doing the building. You don't need to know how to code—the key is learning to state requirements clearly, iterate in small steps to avoid rework, and dodge the spots where UI automation most easily goes wrong.
After Launch: Keep It Running
Take an AI app that runs fine on your own computer and actually put it on the public internet—and keep it alive and stable. This level walks the shortest path to launch, sets up the domain and secrets, builds the basics of monitoring and alerting, and finally teaches you to let an Agent watch production for you, do first-pass triage, and leave the call on the fix to a human.