Seven parts on using AI in programming as an engineer rather than as a consumer. Every piece of code here was actually run, and every output below is real, including the bugs we caught.
📘8 parts🧪all code executed🌓light and dark🌍Arabic and English
Why this guide?
The projects in this track teach you to operate the tools. This guide teaches you to judge them: when to trust, how to verify, and what stays your responsibility.
How to read it
Each part opens with the prerequisites, then the concepts, then the diagrams, then a real exercise with executed output. Do not jump to the code, the judgement is the lesson, not the code.
Cards and questions
Each part ends with flashcards you flip by clicking, multiple-choice questions, and five “check your understanding” questions with hidden answers. Use them before you say “I get it”.
A warning
This guide does not say “the tool is dangerous” or “the tool is magic”. It says one thing: whatever you merge and cannot explain has become your debt, regardless of who wrote it.
The track map
The seven parts are not independent: each builds a habit the next one uses. A dashed line means “this habit carries over there”.
StartKey idea
The order of the parts and the habit each one adds.
Tool names change every year; these six do not. And they are exactly what a senior engineer asks you about when reviewing code your assistant wrote:
StartOutcome
The transferable skills from these eight parts.
The one rule that ties all seven togetherAn assistant generates likely text, not correct text. So every part of this guide is a way of turning “looks right” into evidence: a run, an assertion that fails, official documentation, or a diff review. Whoever owns a measure benefits from the tool; whoever does not is benefited on by the illusion.
Parts added to this track
Part
Topic
You learn
9
AI-assisted code review
A good review request, the checklist, sorting notes, and static tools
10
From a text description to a UI mockup
Describing the screen, iterating, and checking quality
11
An architecture blueprint assistant
Non-functional requirements, components, the diagram, and the decision record
12
Fixing AI-generated code
Bug patterns, test first, and fixing with the smallest change