Google Brings Reusable Gemini Skills to Chrome to Turn Prompts Into One-Click Workflows

Google is giving Chrome a more reusable kind of AI layer. The company has introduced Skills in Chrome, a new Gemini-powered feature that lets users save prompts and run them again across different pages instead of rewriting the same instructions every time.

That may sound small at first glance, but it points to a bigger shift in what browsers are becoming. Chrome is moving beyond a place where users open tabs and manually copy information into AI tools. Google increasingly wants the browser itself to act as a workflow surface where AI actions can persist, travel across pages, and become part of everyday browsing.

From One-Off Prompts to Repeatable Actions

The core idea behind Skills is simple: if a user repeatedly asks Gemini to perform the same kind of task, that instruction can now be saved and reused. Google says users can create a Skill directly from chat history, then trigger it later from Gemini in Chrome by typing a slash or using the add button.

The result is a more agentic browser experience without fully handing over control. A saved Skill can run on the page a user is currently viewing, and in some cases across additional selected tabs, turning common prompt patterns into lightweight repeatable actions. Google gives recipe substitutions as one example, but early usage also points to shopping comparisons, document summaries, and health-related calculations.

Source: Google

Google Is Building a Workflow Layer Inside Chrome

What makes the launch more interesting is that Google is not only shipping custom prompt saving. It is also launching a Skills library with prebuilt workflows covering productivity, shopping, recipes, budgeting, and other common tasks. In effect, Chrome is becoming a distribution layer for reusable AI actions, not just an access point for a chatbot.

That matters because reusable workflows are where AI productivity tools often become genuinely sticky. A one-time summary or rewrite is helpful, but a saved action that works across many pages starts to behave more like software. It gives users a repeatable way to turn Gemini into a browsing assistant tailored to how they actually work.

The company is also keeping a human-in-the-loop model in place. As with other Gemini actions in Chrome, certain higher-consequence tasks will still require explicit confirmation before something like an email send or calendar action goes through. That guardrail is an important signal that Google wants Chrome to feel more capable without making it feel unpredictable.

Source: Google

The Browser Competition Is Becoming More Agentic

Skills also lands in a more competitive browser landscape than Chrome has faced in years. Google’s earlier Gemini integration in Chrome already signaled that the company was taking AI-native browsing seriously, a move TechCrunch previously covered here. Since then, AI browsers and browser-adjacent products from OpenAI, Perplexity, and The Browser Company have made the category feel far less settled.

In that context, Skills looks less like a convenience feature and more like defensive product strategy. If rival browsers are promising agents that browse, reason, and act, Google needs Chrome to offer something more durable than page summaries. Reusable Gemini workflows give it a way to make Chrome smarter without forcing users into an entirely new browsing model.

For now, the rollout is limited to signed-in Chrome desktop users and starts with English (US). But the direction is clear. Google is trying to make the browser a place where AI assistance compounds over time, not just something users summon for isolated tasks. If that sticks, Skills could end up mattering less as a prompt feature than as an early template for what agentic productivity in the browser actually looks like.

Source: Google