Google is testing "Skills" for Gemini in Chrome, which will allow AI in Chrome to perform tasks automatically.

Gemini in Chrome on desktop has been rolling out in the US for a while now.

Right now, Chrome's Gemini acts as an AI “helper” inside the browser.

When you’re on a webpage, you can open Gemini and ask it to explain confusing parts, summarize what you’re reading, or compare info across multiple tabs.

So if you have three tabs open for flights, hotels, and things to do, you can ask Gemini to pull the key details from all of them and turn them into one clear plan.

Google previously confirmed that it plans to turn Gemini into an “agent” inside Chrome over the next few months, and now we're seeing references to "Skills" inside Chrome.

As spotted by Leo, Chrome is adding a new page chrome://skills:

From the "chrome://skills" page, you will be able to add Skills with a name and instructions. Skills allow Gemini to perform specific tasks in Chrome.

This feature is being internally tested in Chrome, and it's unclear when it'll roll out.

Over time, Gemini in Chrome will also help you find pages you visited before using natural questions to work more tightly with Google apps like Calendar, YouTube, and Maps without you needing to switch tabs