Chrome turns your best Gemini prompts into one-click Skills

Repeating the exact same prompt across half a dozen browser tabs is one of the most annoying chores in everyday AI browsing.

Google is rolling out a new feature called Skills in desktop Chrome. Basically, it lets you save your most effective Gemini prompts directly from your chat history and rerun them with a single click anywhere on the web.

Turn prompt history into browser macros

Whenever you craft a prompt you know you will need again—like swapping baking ingredients for vegan alternatives or stripping marketing fluff from product pages—you can now save it as a Skill straight from Gemini's chat panel.

Triggering it later takes almost no effort. You simply open Gemini in Chrome, type a forward slash or tap the plus icon, and select your saved workflow. The model then runs that task against your active tab or across multiple background tabs you choose to bundle together.

Google says early testers leaned heavily on a few core use cases: breaking down macronutrients from online recipes, generating side-by-side spec sheets across competing shopping tabs, and summarizing dense PDF reports.

For anyone who does not feel like prompting from scratch, Google is also bundling a pre-built Skills library. It includes out-of-the-box workflows for cross-referencing gift ideas against budgets and recipient interests, or breaking down chemical ingredient lists. Users can adopt any template and tweak the underlying prompt at will.

On the security side, Google noted that Skills inherit Chrome's layered protections, including automated red-teaming. Any action that alters external services, such as shooting off an email or slotting an event into Google Calendar, will still require manual user confirmation before it executes.

The rollout is landing now on Mac, Windows, and ChromeOS desktop devices, though it is currently restricted to users signed into a Google account with their browser language set to English (US). Saved workflows sync automatically across signed-in machines and can be organized by typing a slash followed by a click on the compass icon.

Why it matters

We think this is a smart, low-friction defensive move from Google.

With challengers like OpenAI's Atlas, Perplexity's Comet, and The Browser Company's Dia trying to reinvent the entire browsing experience around agentic chat, Google cannot afford to let Chrome feel static. Instead of forcing users into an unfamiliar interface, Skills quietly turns the classic tab strip into an automated workspace.

For everyday users and knowledge workers, it eliminates the messy copy-paste shuffle. You no longer need third-party extensions, userscripts, or separate note files just to hold your favorite prompt templates. Having slash-command shortcuts native to the browser makes lightweight AI automation accessible to people who would never touch an API or workflow builder.

The limitation, however, is that this is still prompt chaining rather than true autonomous web agents. It relies strictly on in-context Gemini analysis and basic tab scraping, and it remains locked inside Google's ecosystem. Whether slash-command macros are enough to stop power users from migrating to dedicated AI browsers remains an open question, but for the average Chrome user, it is immediately useful today.