GitHub Copilot App for Beginners: Using Diff, Terminal, and Browser

When an agent makes changes to your code, you want to preview it, run it, and see what changed. Great news: you can now do all three without leaving the GitHub Copilot app.

Previously, performing these three tasks meant having to switch between an editor, a terminal window, and a web browser. But when these steps are side by side as built-in panels in the Copilot app, checking your agent’s performance becomes easier.

Let’s look at each of the app’s panels and how we’ll use them to complete the AI ​​coding loop.

Using the Differences Panel to View Changes

The diff panel (diff is a before and after comparison) shows exactly which rows have been added, deleted, or changed, with additions highlighted in green and deletions in red.

This gives you complete clarity and you can choose what happens. You can accept the changes, leave comments, or ask the co-pilot to make changes. You are in complete control of the situation and make the final decisions.

Running commands in the terminal panel

Reading code is good, but doing it is even better. You can run commands directly within your session from the Copilot app’s terminal panel. And if it looks scary, don’t worry. Basically you just run your own project commands and read the results.

You can run the code manually or configure it as a script (available via Run button). Imagine you are working on a website. Here’s an example of how it works:

• Add dev server a script that opens the client folder and then runs npm run dev.
• Click Run start a server for the site.

You can open multiple terminal windows at the same time so you can switch between them and continue executing commands.

For anything that has a user interface, the browser bar completes the loop. And for the example of our website, this panel means that you can open it and test a new feature as if you were using the site.

If you want to continue iterating you can use Select and polish a tool for selecting an element and configuring it using an agent.

And to see what has been fixed, you can run the command dev server script again after making changes.

Completing the AI ​​Code Review Cycle

Now you’ve inspected the difference, run the project in the terminal, viewed it in the browser, and iterated to get a working function. Once you have what you want, you can commit the change directly from the Copilot app and create a pull request. Completing this cycle from one place means you won’t have to jump between tabs, switch apps, or lose your place.

Parallel review, launch, and preview is what makes changes made by an agent safe rather than scary because you can prove that what you put together works.

Take it with you

These built-in panels answer three important questions you should always ask before accepting an agent job.

• What has changed?
• Is it working?
• And does it In fact Job?

Reviewing this list before accepting agent-generated code will help you understand what has changed and whether it works the way you want while maintaining control. ✨

Posted by

Senior AI Developer Tools Specialist