Google Chrome Introduces Skills to Automate AI Prompts with One Click
The integration of generative AI within browsers is undergoing a fundamental transformation, shifting from a single interaction model to that of structured workflows. Google also wants to jump on this bandwagon and has announced the debut of Skills in Chrome, a feature designed to eliminate the operational redundancy typical of those using Gemini for repetitive tasks. Until now, executing a specific operation across different web pages required manually entering the same prompt for each session. With the introduction of Skills, Mountain View allows users to transform any effective prompt into a reusable macro, instantly callable through the Gemini interface integrated into the browser.
The architecture of Skills directly leverages the interface of Gemini in Chrome. Users can save an AI command directly from their chat history when they find a satisfactory output. To call a saved Skill, simply type the command forward slash ( / ) in the input bar or click the “+” button. This quick recall system introduces a capability for multi-tab processing, allowing users to select different open tabs and run the Skill simultaneously across all of them. For example, this way, a user can compare the technical specifications of different products or summarize multiple documents without having to physically navigate between the tabs.
How to Activate and Manage Skills in Gemini (Not Yet Available in Italy)
In addition to manual creation, Google has made available a Skills library, a collection of ready-to-use workflows that cover the most common needs of daily browsing. Among the cited examples are tools for instantaneous calculation of macronutrients based on online recipes, product comparison systems for shopping, and scanning tools for lengthy documents.
The logic of the system is not closed: every Skill, whether created by the user or taken from the library, can be remixed. This means that the prompt underlying the Skill can be edited and refined at any time to better fit specific needs. Management of the entire ecosystem occurs through a dedicated interface, accessible by clicking the compass icon within Gemini, which ensures synchronization across all desktop devices the user is logged into.
A critical aspect relates to delegating complex actions to artificial intelligence. Google has implemented various levels of protection to ensure that automation does not escape the user’s control. Skills that require interactions with productivity apps, such as adding events to the calendar or sending emails, will always require explicit confirmation before proceeding. From an infrastructural perspective, the function benefits from Chrome's automated red-teaming system and automatic update cycles aimed at preventing exploits related to the injection of malicious prompts or unauthorized access to browsing data. The rollout of the feature is already underway for Chrome users on desktop systems, but currently, it is only available if the browser is set to English-US language.