Open curriculum · MIT licensed
You don’t just learn AI.
You build it. By hand.
523 lessons across 20 phases in Python, TypeScript, Rust and Julia — and every lesson ships a reusable artifact: a prompt, a skill, an agent, an MCP server. Free, open source, MIT.
Use every lesson the same way
step 5 is a gate-
01
Read
Open the lesson and explain the core idea in your own words.
-
02
Type and build
Type the important code instead of treating the code block as decoration.
-
03
Run
Run the lesson command from the repository root.
-
04
Keep evidence
The command, working directory, exit code, meaningful output, and the artifact you produced.
-
05
Continue — only when
You can explain the output and make one small change without guessing.
Four steps, then a condition. Most curricula end at step three; this one refuses to advance you until the last line is true.
What a lesson leaves behind
the project’s own four artifact typesWhy evidence is step four, not an afterthought
The first thing the quickstart asks you to do is run a dependency-free lesson and save that terminal output as your first evidence. The preflight separates requirements needed now from tools needed later, and every required failure includes the detected reason and a corrective command — which is the difference between a course and a harness.
Lesson, phase, hour and language counts are the project’s published figures, reproduced as stated. No completion rate, reader count or outcome claim is reproduced here.