Introduction
Written by the author, AI-editedAI-generated, author-verified How this module was made
In the last lesson, you reviewed the Tokens Dashboard app manually the way you would review any teammate’s pull request. You found four things worth fixing: a naming inconsistency, a design inconsistency, a force-unwrap, and duplicated business logic between two view models.
Now when you move on to the AI as a second reviewer, the questions are: what does AI find when you ask it to review the same app? Does it find what you discovered? Does it find things you missed? And how do you know which of its suggestions to actually trust?
That is what this lesson is about. You’ll run two review prompts against the same app. First, a completely open review with no guidance, and then a structured review with the same seven-category checklist you used by hand, and compare what each one surfaces.
By the end of this lesson, you’ll be able to:
- Use AI as a second reviewer rather than the primary one.
- Compare unguided and structured AI review output against your own manual findings.
- Prompt AI from different perspectives to surface different kinds of issues.
- Decide whether to accept, reject, or investigate further for each AI suggestion.