Google is adding a new feature to Chrome: “Skills”. These tools allow you to record Gemini prompts and automatically relaunch them on any web page.
The integration of artificial intelligence into browsers continues to intensify. After introducing its Gemini assistant into Chrome, Google is taking a new step with a feature called “Skills”.
The goal is to allow users to save frequently used prompts and then instantly rerun them on different web pages. An evolution which aims to transform certain uses of AI into real automations directly in the browser…
Reusable Gemini prompts in one click
Officially announced on The Keyword website, with Skills, Google seeks to respond to a recurring problem in the use of AI: the repetition of the same queries. Until now, to perform the same task on several pages, for example analyzing a product or comparing information, you had to enter the prompt each time.
The new feature now allows you to record a prompt directly from the chat history in the Gemini side panel in Chrome. Once saved, it becomes a Skill, that is to say a reusable command.
To trigger it, the user can simply type “/” followed by the name of the Skill or click on the “+” button in the interface. The command then executes on the active page, but also on several selected tabs if the task requires it.
This operation opens the door to several practical uses. For example, it becomes possible to automatically compare technical sheets opened in several tabs, to analyze the composition of a product consulted online or to calculate the nutritional information of a recipe.
Saved Skills are also synced to the user’s Google account. They can therefore be found on all computers using Chrome with the same account.
A library of Skills already available
The Skills library is already very complete – Source: Google
To support the launch of this feature, Google already offers a library of ready-to-use Skills. These cover common tasks like finding a gift within a given budget, analyzing food ingredients or generating comparisons between several products.
Each Skill can be modified and personalized to meet the user’s needs. Google also specifies that these automations benefit from the same security mechanisms as the classic Gemini prompts. In some cases, confirmation will be requested before performing a binding action, such as sending an email or adding an event to a calendar.
Note that for the moment, the functionality is being deployed on the desktop versions of Chrome, and remains limited to users whose browser is configured in American English.
Finally, Google has not yet communicated a timetable for availability in other languages, which means that French-speaking users will still have to wait before being able to test these new automation tools…
