On April 14, 2026, Google began rolling out a Chrome desktop feature called Skills: save a prompt you have used in Gemini in Chrome, then rerun it on any web page with one click instead of retyping it. The rollout covers Mac, Windows, and ChromeOS starting that day, gated on two conditions — the browser language must be set to US English, and you must be signed into a Google account.
The feature looks small, but the direction is clear. The AI browser assistant race has moved past “can it answer questions about this page” and onto “can it turn your repeated operations into reusable assets.” Skills is Google’s first formal answer to that question.
What Skills Is: Prompts Become Shortcuts
The mechanics are simple. From your Gemini in Chrome chat history, you save a prompt that worked well as a Skill. Invoke it later by typing a forward slash (/) or clicking the plus (+) button. The Skill runs against the page you are viewing, and you can select additional tabs to include. Saved Skills stay editable, are managed through a compass icon in the slash menu, and sync across every signed-in Chrome desktop device.
Google is also shipping a Skills library of ready-made workflows, with categories spanning productivity, shopping, recipes, and budgeting. Cited tester use cases include calculating protein macros from a recipe page, comparing product specs side by side across tabs, and extracting key information from long documents. Library Skills can be added and then edited like your own — Google explicitly treats the library as a starting point, not a fixed catalog, and the testers it observed tended to customize what they adopted. The library is also the obvious surface for future ecosystem moves: a format that third parties could publish into would turn saved prompts into shareable objects, the way browser extensions became distributable behavior. Google has not announced anything like that yet, but the data model — a named, editable, invocable prompt bound to page context — is the right primitive for it.
Why Now: The AI Browser Race
TechCrunch places Skills in its competitive context. Gemini in Chrome already handled page questions, summaries, and task execution, and its arrival coincided with a wave of AI-native browsers: OpenAI’s Atlas, Perplexity’s Comet, and The Browser Company’s Dia. Now that everyone can “let AI read your page,” Google’s differentiator is distribution — Skills lands inside the browser billions of people already use, rather than asking them to switch. It is the same logic running through consumer AI in 2026: the channel is harder to copy than the feature.
The Safety Model: Confirmation Before Action
Skills inherits the safeguards already applied to prompts in Gemini in Chrome, and it asks for explicit confirmation before consequential actions, such as sending an email or adding a calendar event. Google also notes that Skills benefits from Chrome’s layered protections, including automated red-teaming and auto-updates. For product teams, the confirm-before-acting design is the detail worth studying: once a prompt becomes a shortcut you rerun regularly, the cost of an accidental trigger grows. The confirmation step is necessary friction, not leftover UI.
What It Suggests for Product Builders
Three observations. First, prompts are becoming user-level macros. When someone saves a prompt for reuse, it has crossed from a one-off conversation into a stable workflow — a genuine signal of AI product stickiness, and one that is measurable: a growing library of saved Skills is retention data a product team can actually watch. Second, the library-plus-customization combo is a classic distribution play: prebuilt Skills lower the entry barrier, while editing keeps power users invested. The same pattern governed spreadsheet templates and text-editor snippets; prompts are simply the newest substrate. Third, the current limits — desktop only, US English only — show Google buying safety headroom with a conservative rollout. The cross-device sync already implies server-side storage tied to a Google account, which makes each added language and surface area a privacy surface too. Mobile support and additional languages are the next things to watch, and the pace of that expansion will say a lot about how the feature is performing internally.
Sources
- Google adds AI Skills to Chrome to help you save favorite workflows — TechCrunch
- Skills in Chrome — Google Blog
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.