Gemini in Chrome is getting “Skills” as it moves toward becoming a full AI agent

Google has spent the last year turning Gemini into a versatile helper within the Chrome browser, but the platform is now preparing for a significant shift in how it interacts with the web. Rather than just acting as a sidebar assistant that summarizes text or compares prices across tabs, Gemini is being upgraded into a proactive “agent” capable of performing complex tasks on your behalf.

Teaching Gemini new tricks

Recent discoveries in the Chromium Gerrit point to a new feature called “Skills,” which appears to be the foundation for this transition. A hidden internal page, chrome://skills, has been spotted in testing, revealing a dedicated interface where users will eventually be able to define specific capabilities for the AI.

Gemini in Chrome is getting "Skills." This feature allows the AI to perform specialized tasks automatically. You have to add a skill on the "chrome://skills" page with a name and instructions. Then open the side panel on a site to let Gemini run the skill. The page is empty atm: pic.twitter.com/a7ruaI6tFp

— Leopeva64 (@Leopeva64) January 16, 2026

Based on early looks at the code, users will be able to give these skills a name and a custom set of instructions. This effectively allows you to “teach” Gemini how to handle specific or repetitive workflows within the browser. Instead of Gemini just being a generic chatbot, it should become a personalized tool tuned to your specific needs.

From passive helper to active agent

This development aligns with Google’s previously stated goals to evolve Gemini from a conversational bot into a functional agent. Up until now, features like “Help me read” or tab comparisons have been relatively passive: you provide the context, and Gemini provides the output.

With the introduction of Skills, the AI moves toward execution. It will eventually interact with the browser’s architecture to get things done without the user needing to navigate every single step. This is the bridge between asking a question and having the AI actually perform the work for you.

Deep integration across the Google ecosystem

The broader vision for this integration involves Gemini working more tightly with the rest of the Google ecosystem directly from the omnibox or sidebar. Over the coming months, we expect to see Gemini gain the ability to search through your browsing history using natural language and interact with apps like Google Calendar, YouTube, and Maps without requiring you to switch tabs.

For example, instead of manually finding a confirmation email and then opening Calendar to add an event, a “Skill” could allow Gemini to identify those details on a webpage and handle the scheduling automatically. While these AI Skills are currently under internal testing, their presence suggests that the era of the more-capable AI agent right in the browser is soon to be upon us.