Alfredo Juan
Back to the workbench

02Tool

VibeExplainer101

To understand what was built

VibeExplainer101 organizes a project into modules, components and files. It is designed to help people who build with AI investigate the result, following connections between an explanation and the code supporting it.

Recipe form✓ verified API✓ verified Storage✓ verified Accounts~ inferred server/db.ts verified in the codepartly inferred

Why I built it

I started building software with AI back when that still meant copying and pasting code, function by function. Today I can build much bigger projects, and it is not always easy to understand everything that has been built. Sometimes the project works, I look at that pile of files and think: draw me a picture. VibeExplainer101 came out of that need: to have a map of the software, start from a simple view and be able to go into the details. It is a difficulty of my own that I also see in other people building with AI.

A small example

“Recipe box” is a fictional recipe app, used as an example in the project. The diagram below was drawn from it for this page; it is not a VibeExplainer101 screenshot. It shows the difference between a part examined in detail and a relationship that is still partly inferred.

Recipe box

RecipesRecipe formVerified
APIReceives editsVerified
StorageRecipesVerified
Accounts↝ Partly inferred

An explanation you can examine

The project links explanations to files and code excerpts. It also distinguishes verified findings from partially inferred ones. This indication helps investigate a claim; it does not, by itself, guarantee that the explanation is correct.

In development

The interface and scope are still evolving. For now, this page shows only the concept.