Google adds Skills to Chrome: save AI prompts and run them in one click

Google has launched a new Chrome feature called Skills, which turns effective Gemini prompts into reusable commands. Instead of retyping the same prompt on every site, you can save it and run it in one click on the current page or across several tabs at once.

How Skills work

Skills are built into Gemini in Chrome and designed for repeatable tasks: quickly comparing products, pulling facts from a long document, checking a product's ingredients, or adapting a recipe. When a user gets a useful result, they can save the prompt straight from the chat history and later call it up with the `/` character or the `+` button in the browser's side panel.

The point of the update is not a new model but packaging a prompt as a permanent tool. The prompt stays editable: you can rewrite it, refine it for your task, or create several variants for different cases. Google specifically stresses that a saved Skill runs on the current page and can take other selected tabs into account if the task requires comparison or a summary across several sources at once.

For Chrome, this is also a product move in the fight to make the browser the main AI surface. Rather than a separate app, Google is building repeatable workflows directly into web browsing: where users read, compare, buy and make decisions. Other players are also becoming more active in this space, so betting on fast, one-step AI actions inside the browser looks pragmatic.

Ready-made workflows

Alongside user-created Skills, Google has opened a library of ready-made templates for common tasks. This should lower the barrier for people who don't want to come up with effective wording themselves and test it in Gemini. You can add a template you find to your collection and then, if you wish, edit it to fit your own way of working.

• Comparing product specs across several tabs
• Finding key points in long documents
• Counting nutrients and macros in recipes
• Analyzing a product's ingredients before buying
• Choosing a gift based on budget and the person's interests

According to Google, early testers used Skills most often for shopping, productivity, and health and nutrition tasks. This shows where Chrome is steering its built-in AI: not toward a separate chat window for experiments, but toward regular browser actions you want to repeat every day without copy-pasting the same instructions.

Launch and limitations

Google began rolling out Skills on April 14, 2026, for Gemini in Chrome on desktop. At launch, the feature is available on Mac, Windows and ChromeOS if the browser language is set to English (US). Saved Skills sync across desktop devices where the user is signed in to their Chrome account, so familiar AI workflows can move between a work and a home computer without manual setup.

Google also emphasizes user control. If a Skill leads to an action beyond a simple answer, such as adding a calendar event or sending an email, the browser must ask for confirmation. The company says the feature uses the same safeguards as Gemini in Chrome, including Chrome's built-in security measures, automatic updates and internal red teaming before release.

What this means

For Google, this is a way to make Gemini in the browser more useful without grand agentic promises: the company is selling not an "autonomous assistant" but a set of repeatable mini-workflows on top of the web. For users and teams, it is an important shift: your best prompt stops living in notes or an old chat and becomes a button you can run as needed on any suitable page.