Rethinking the Review Process in Chat

When we talk about ChatGPT & MCP Apps, we often focus on adding new capabilities like searching the web or generating images. But some of the most useful apps do not add new data. They change how we consume what is already there.

ChatGPT is a world class writer. I use it constantly to refine emails, write blog posts, and polish contracts. The problem I kept running into had nothing to do with output quality. It was the output visual.

When ChatGPT gives me a revised version of something I wrote, the original version is quickly pushed off screen. If I want to understand what actually changed, I find myself scrolling up and down, trying to remember what the earlier version looked like. It turns into a frustrating game of spot the difference.

Even when I ask ChatGPT directly what it changed, the response is usually an explanation in plain text, I need to go find where they are. Technically helpful, but still very hard to track. I do not want a paragraph describing edits. I want to see them.

A Simple App That Visualizes Change

To solve this, I built a Text Compare app.

It does not change how ChatGPT writes. It changes how I review what it wrote.

Instead of returning another wall of text, the app opens a dedicated UI that shows both versions side by side, with additions and deletions clearly highlighted. Red for removals. Green for additions.

I also added full screen support, which gives me enough space to actually review longer edits without feeling cramped.

Building the App with a Single Prompt

This app was built using Fractal, which is the simplest way I have found to build ChatGPT & MCP Apps. Everything starts with a single structured prompt.

Below is the exact prompt used to create the app.

Create a Text compare app.

The app is invoked when the user wants to compare 2 texts, like between 2 options ChatGPT gives, or a version ChatGPT edited with the original version.

The app shows the difference between 2 texts. The app shows 2 texts side by side on desktop and supports full screen mode.

Breaking Down the Anatomy of the Prompt

This prompt follows the same four part structure I use for all high utility ChatGPT & MCP Apps built with Fractal.

1. A General Description of What You Want to Build

"Create a Text compare app."

This sets the intent. It tells the system we are building a comparing text app.

2. When the App Should Be Called

The app is invoked only when I want to compare two texts, whether that is two options ChatGPT generated or an edited version versus the original. This helps ChatGPT understand when to show the app.

3. What Input the App Receives

The app receives two text strings that need to be compared.

4. What the App Does with That Input

The app renders both texts side by side, highlights the differences, and supports full screen mode.

The Takeaway

The Text Compare app is simple, but it removes a lot of friction from my workflow. By solving a basic UI limitation, ChatGPT becomes a much better editing tool.

This is also a good example of why Fractal works so well for building ChatGPT & MCP Apps. You do not need a long spec or complex logic. If you can clearly describe when an app should appear and what it should do, you can build something immediately useful.

If you ever notice yourself frustrated by the way ChatGPT presents something, that is usually a sign an app should exist.

Want to see more examples of how small UI changes can dramatically improve your workflow? I have a YouTube playlist that covers how to build ChatGPT & MCP Apps with Fractal here: Watch the YouTube Playlist

Ready to Build Your ChatGPT & MCP Apps?

Start building with Fractal today and get your app discovered by millions of ChatGPT users.