Google is baking a new layer of AI right into Chrome, and this time it is not just another sidebar chatbot – it is something you can turn into muscle memory. The new “Skills in Chrome” feature lets you save your best Gemini prompts as one-click tools that run on whatever page – or set of tabs – you are looking at, turning casual AI tinkering into reusable workflows.

At its core, Skills are Google’s answer to a very common AI pain point: repetition. If you have ever found yourself typing the same “make this recipe high-protein,” “summarize this article,” or “compare these two products” prompt over and over again, Skills basically say: save that once, and never retype it again. Inside Gemini in Chrome’s side panel, any prompt you run can be saved directly from your chat history as a Skill – a named, reusable mini-tool. Next time, you just type a forward slash (/) or hit the little plus button in the Gemini panel, pick your Skill, and it runs on the page you are on, as well as any other tabs you choose.

That multi-tab part is worth pausing on, because it is one of the capabilities that makes this feel more like native browser automation than a generic AI overlay. Instead of copying links into a chatbot or juggling multiple windows, you can select several open tabs, fire a single Skill, and have Gemini scan everything in one shot – useful for things like comparing specs across shopping sites, pulling key points from a cluster of research articles, or skimming a stack of docs for deadlines and action items. It is the kind of workflow power people usually hack together with extensions or scripts; now it is shipping in Chrome itself, no extra install required.

Google is not leaving you on your own to design all of this from scratch either. Alongside the feature, there is a built-in Skills library – think of it as a starter kit of ready-made workflows – that you can browse from a dedicated page (chrome://skills/browse) or via the Gemini panel. These premade Skills cover common use cases like breaking down the ingredients of a product you are looking at, cross-checking a gift idea against your budget and the recipient’s interests, maximizing protein in recipes, summarizing YouTube videos, or scanning long documents for key information. You can add any of them to your own collection with a click and then edit the underlying prompt to match your style or your workflow.

If you are the kind of power user who enjoys tweaking, Google has built in some nice touches. When you save a Skill, you can give it a custom name and even attach an emoji so it is easier to spot in a growing list of tools. Gemini will also nudge you to “turn this into a Skill” after certain chats, which is a subtle way of teaching people that any good prompt is a candidate to be turned into a reusable workflow. All your saved Skills sync across signed-in Chrome desktop devices, so whatever you build on your work laptop will show up on your home PC as well, as long as you are logged into the same Google account.

On the rollout side, Google is starting with a pretty specific slice of users: desktop Chrome on Windows, macOS and ChromeOS, with Gemini enabled and the browser language set to English. You do not need a paid AI subscription – the company has been explicit that Skills ride on top of the free Gemini in Chrome integration, not a premium plan. As with most big Chrome features, the rollout is staged, so even if you meet the requirements, it might take a few days before Skills shows up in your browser.

Under the hood, Skills are very much part of the broader Gemini-in-Chrome push: the side panel where you chat with Gemini, the ability to ask questions about the page you are on, the way the AI can understand the context of what you are viewing. The difference is that Skills are designed for repeatability. Instead of thinking of AI as a one-off assistant you query and forget, Google wants you to gradually build a personal toolbox – “my recipe fixer,” “my spec comparer,” “my legalese translator” – and then lean on those tools every time you browse.

Given how sensitive AI-in-the-browser can feel, Google is very deliberately stressing the privacy and security story. The company says Skills are built on Chrome’s existing security foundation and use the same safeguards as regular Gemini prompts, including things like automated red-teaming and Chrome’s constant auto-update system to patch issues quickly. Skills that touch personal or sensitive actions – adding calendar events, drafting emails, that sort of thing – will always ask for your explicit confirmation before doing anything, rather than silently firing off changes in the background. For people coming from the enterprise side, a lot of this builds on work Google has already done to make Gemini in Chrome manageable in corporate environments, with admin controls and data protection policies layered on top.

From a user-experience perspective, Skills also quietly shift the mental model of what “AI in the browser” is supposed to be. Instead of trying to be a destination – a separate AI site you visit – Gemini is starting to look more like a system capability, closer to how we think about bookmarks, extensions, or even keyboard shortcuts. You are no longer just chatting with a model; you are wiring the model into the things you do all the time: shopping comparisons, learning, content drafting, research, planning. The fact that Skills can run against multiple tabs at once in a native way is something competing browser-based AI tools mostly do not offer yet, which gives Chrome an interesting edge for workflow-heavy users.

Of course, this is still early days, and we do not have a full picture of how ordinary users will adopt Skills versus just continuing to type ad-hoc prompts when they need something. There is also the broader question of how this blends with the rest of Chrome’s AI experiments, like vertical tabs, new reading modes, and other Gemini-powered features the browser has been steadily rolling out over the past year. But if you zoom out, Skills fits into a clear pattern: Google is trying to make the browser a place where your AI workflows live natively, instead of being something you bolt on with third-party extensions.

If you are a Chrome user in the US with Gemini already turned on, the easiest way to think about Skills is this: the next time you catch yourself copying the same prompt into Gemini more than once, consider saving it. Name it, give it an emoji, maybe tweak the wording a bit, and suddenly you have your own one-click tool that follows you around the web. Over time, that collection of Skills could end up being as personal and useful as your bookmarks bar – just a lot smarter.