How to Use Google Chrome's New AI-Powered 'Skills'
Google launched a new AI feature called Skills inside Chrome on September 18, 2025, giving users access to more than 50 preset AI prompts they can trigger with a forward slash in the Gemini sidebar. This matters because it signals a broader shift from open-ended chatbot interface...
Google just made its Chrome browser feel a lot more like a personal assistant and a lot less like a search box. According to Wired's coverage of the launch, Google rolled out a feature called Skills on September 18, 2025, embedding repeatable AI prompts directly into the Gemini sidebar that lives on the right side of the Chrome browser. The feature lets users pick from a library of over 50 prebuilt prompt templates, or build their own, and run them against whatever webpage they happen to be looking at. No prompt engineering required.
Why This Matters
Google is not experimenting here. This is a direct play to make Chrome sticky in a world where browsers are increasingly competing on AI capability rather than rendering speed or extension libraries. With more than 50 preset Skills available at launch, Google is betting that lowering the friction to AI adoption, by removing the blank prompt box entirely, will drive daily active use of Gemini inside Chrome in a way that a generic chatbot sidebar never could. Microsoft has been doing something similar with Copilot in Edge, but Google's approach of tying Skills to keyboard shortcuts and a slash-command interface borrows directly from the playbook that made Slack slash commands and Notion templates popular with productivity-focused users.
Daily briefing from 50+ sources. Free, 5-minute read.
The Full Story
Skills work through the Gemini sidebar that Google added to Chrome earlier in 2025. To use them, you click the "Ask Gemini" sparkle icon in the upper-right corner of the browser window, which opens the sidebar. Once inside, typing a forward slash in the prompt box pulls up the Skills library, where you can scroll through prebuilt options or search for one that fits what you are trying to do. After selecting a Skill, Gemini analyzes the content of the browser tabs you have shared with it and returns a response shaped by the specific instructions baked into that Skill.
The prebuilt Skills cover a wide range of everyday tasks. The "Protein Maximizer" Skill, for example, tells Gemini to scan the recipe on your current webpage, identify every ingredient, estimate its protein content, suggest substitutions or additions to increase total protein, and then output a revised recipe that shows protein content per ingredient and total protein per serving. That is a genuinely complex prompt that most users would never think to write themselves. Google has done the prompt-crafting work for them.
The Skills library at launch covers areas including YouTube video summarization, recipe modification, and job listing evaluation. Users can also create custom Skills using Gemini directly in the browser, which opens the door for developers and productivity-focused users to build their own repeatable workflows without leaving Chrome. This is where the feature could get interesting for teams that want consistent AI-assisted research or content analysis processes across a shared browser environment.
For users who want nothing to do with this, Google built an opt-out. Going into Chrome Settings, then the AI Innovations tab, then the Gemini in Chrome section, and toggling off the top switch removes the Ask Gemini button from the browser entirely. That is a cleaner opt-out path than most AI browser features offer, and it matters because not every Chrome user wants a generative AI assistant watching their tabs.
Wired also notes that this Skills launch is part of a broader pattern. Google added Gemini to the Chrome sidebar earlier in 2025, and the company has been testing features that let Gemini actually click and browse the web on a user's behalf. Skills fits into that arc as a more controlled, template-driven version of that autonomous browsing concept, one where the AI does exactly what the preset instructs rather than interpreting open-ended instructions.
Key Details
- Google launched Skills on September 18, 2025, as part of a broader set of new AI features for Chrome.
- The Skills library contains more than 50 prebuilt prompt templates at launch.
- Users access Skills by typing a forward slash in the Gemini sidebar prompt box.
- Example Skills include YouTube video summarization, a Protein Maximizer recipe tool, and a job listing evaluator.
- The Gemini sidebar was first added to Chrome earlier in 2025, establishing the foundation Skills now builds .
- Users can disable the Ask Gemini button entirely through Settings, then AI Innovations, then Gemini in Chrome.
- Custom Skill creation is available to any user directly through the Gemini interface in Chrome.
What's Next
Google has already signaled that Skills is one component of a September 2025 rollout that included 10 new AI updates for Chrome, so expect additional preset Skills to appear in the library over the coming months as Google identifies which use cases drive the most engagement. The more interesting development to watch is whether third-party developers gain the ability to publish Skills to the library the way they publish Chrome extensions today. If that happens, the Skills library could grow from 50 templates to thousands within a year.
How This Compares
Microsoft's Copilot integration in Edge is the most direct comparison, and Skills actually has a more intuitive access pattern. Edge Copilot requires users to open a sidebar and formulate their own requests, while Google's slash-command trigger and prebuilt library removes that cognitive step entirely. It is a small difference in interface design but a significant one for adoption. Slack proved years ago that slash commands lower the barrier to feature discovery, and Google appears to have studied that lesson carefully.
The Opera Neon browser, which Wired references as a smaller Chrome competitor, has also been experimenting with integrated AI prompts, but Opera does not have Google's distribution. Chrome holds a dominant share of the browser market, which means Skills could reach hundreds of millions of users without requiring anyone to download a new application. That scale advantage is something Microsoft with Edge and Opera without it simply cannot match.
Apple's approach to on-device AI in Safari and across iOS goes in a fundamentally different direction. Apple emphasizes local processing and privacy-first design, which limits what its AI can do with real-time web content compared to what Gemini can do when given access to your open tabs. Google's Skills feature is more powerful in the tasks it can perform, but it also requires sharing browsing context with Google's servers, a tradeoff that privacy-conscious users should weigh before enabling it. These are two legitimate but genuinely different philosophies about what AI in a browser should be allowed to .
FAQ
Q: Do I need a Google account to use Chrome Skills? A: Yes, Skills run through the Gemini sidebar, which requires you to be signed into a Google account in Chrome. The feature ties into your Gemini access, so you will need to have Gemini enabled for your account before the Skills library becomes available.
Q: Can Chrome Skills see everything in my browser? A: Gemini analyzes content from the browser tabs you choose to share with it, not your entire browsing history automatically. You control which tabs the sidebar has access to, though Google does process that shared content on its servers, so it is worth reviewing Google's privacy policy if that concerns you.
Q: How do I create a custom Skill in Chrome? A: Open the Gemini sidebar by clicking the Ask Gemini sparkle icon in the upper-right corner of Chrome, then follow the prompts to build a custom Skill using natural language instructions. You define what Gemini should do, what information it should look for, and how it should format its output, and Chrome saves it for reuse with a keyboard shortcut.
Google's Skills feature is a practical bet that most users will adopt AI tools faster when someone else has already done the hard work of figuring out the right prompt. Whether the library grows into something developers can contribute to, or stays a curated Google product, will determine just how useful it becomes over the next year. Subscribe to the AI Agents Daily weekly newsletter for daily updates on AI agents, tools, and automation.
Get stories like this daily
Free briefing. Curated from 50+ sources. 5-minute read every morning.
