Google adds Skills to Chrome, letting Gemini reuse prompts across sites

Google is turning Chrome into more than a window to the web. With Skills, announced April 14, the company is giving Gemini users one-click workflows they can save, rename and replay across sites, a move that pushes Chrome deeper into the AI browser fight and raises the stakes around lock-in, data access and prompt portability.

Skills work by letting users launch saved instructions from Gemini in Chrome with a forward slash or a plus button. A recipe shopper could save a prompt that asks Gemini to suggest vegan substitutions, while a product researcher could reuse the same workflow to compare items across multiple tabs. Google said saved Skills can run on the current page and on other selected tabs, which is the key strategic shift: the browser is no longer just surfacing answers, it is preserving user intent inside Chrome.

Google is also building a Skills library at chrome://skills/browse with prewritten workflows in categories including Learning, Research, Shopping and Writing. Users can save a Skill from chat history, rename it, assign an emoji and edit it later. Google said early testers were already using similar workflows for health and wellness, shopping and document scanning, evidence that the feature is meant to reach beyond technical power users and into ordinary browsing habits.

The company is also drawing a line around safety. Skills will ask for confirmation before sensitive actions such as sending an email or adding a calendar event, and Google said they use the same security and privacy safeguards as Gemini in Chrome. That matters because the more tasks Chrome handles across tabs, the more personal context Google’s assistant can see. In the browser wars, that context is the prize.

The rollout starts with Chrome desktop users who are signed in, in English, in the United States. It lands as Google faces a tighter field of AI-first rivals, including OpenAI’s Atlas, Perplexity’s Comet and The Browser Company’s Dia. Google first framed Gemini in Chrome in September 2025 as an AI browsing assistant that could answer questions across multiple tabs, and its Help documentation says it can already handle multi-step tasks such as comparing products, searching for deals, booking travel and making restaurant reservations. The auto-browse feature is being released gradually and is not available in Live chats or on iPhone.

The business case is clear. If Skills become part of users’ daily routines, Chrome becomes a persistent AI workspace, not just a browser. If they do not, the feature may still serve Google’s broader goal: strengthening Gemini’s distribution advantage at the center of the web.