What Gemini Spark Is and Who Can Use It
Gemini Spark is a personal AI agent from Google that runs continuously in the background, connects to your emails, calendar, and documents, and carries out digital tasks on your behalf with semi-autonomous decision-making while still seeking your approval for major actions. At launch, Gemini Spark is limited to people on the Google AI Ultra subscription, a premium plan that now starts at USD 99.99 (approx. RM460) per month. Ultra subscribers can access the Gemini Spark AI agent on Android, iOS, and the web, where it lives alongside the standard Gemini chat view. Google describes Spark as a way to “help you handle your digital life” by staying active even when your phone or laptop is turned off, turning Gemini from a chat-style personal AI assistant into a more persistent layer of AI workplace automation across Google Workspace.
Tasks, Skills, and Schedules: How Spark Works in Practice
Gemini Spark is built around three pillars—Tasks, Skills, and Schedules—that define how it works with your Gmail, Calendar, Docs, Sheets, and Slides data. Tasks are one-off jobs, such as “Help me find and track interior design internships in New Orleans for this summer,” where the agent searches, organizes results, and updates your files or to-do lists. Skills are reusable behaviors: you might ask Spark to read your last 50 sent emails, extract your writing style, and save it as a “ghostwriter” skill that runs every time it drafts email on your behalf. Schedules add time- or condition-based triggers, like scanning your inbox every Monday at 9:00 AM, summarizing important messages, turning them into a prioritized to-do list, and placing deep-work blocks directly on your calendar. Together, these features aim to shift Gemini Spark from reactive chatbot to proactive workflow partner.
Hands-On Reality: Powerful, Persistent, and Sometimes Tone-Deaf
Early hands-on reports show that the Gemini Spark AI agent can manage complex chains of activity across your calendar, inbox, and documents, but it can still misread subtle context in messy human situations. Because it runs 24/7 once enabled, Spark is good at slow-burn tasks like ongoing inbox triage, long-range planning, or maintaining shared documents without constant prompting. However, tests have highlighted awkward moments where it applies rules too literally, such as over-sanitizing personal messages or misjudging relationship dynamics in email drafts. Spark does check in before taking major actions, which helps keep control in your hands, but users expecting flawless judgment may be disappointed. At its best, it behaves like a tireless personal AI assistant that keeps your digital house in order; at its worst, it feels like an over-eager intern who misses nuance.
Beyond the App: Google Workspace CLI and Agent Workflows
For power users, Gemini Spark sits inside a larger shift toward AI workplace automation around Google Workspace. The new Google Workspace CLI, known as gws, gives a single command-line entry point for Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more. Written in Rust and licensed under Apache 2.0, it reads Google’s Discovery Service at runtime to build commands dynamically instead of shipping a fixed command list. That means when an API changes, the CLI updates its surface without a new release. According to Google Cloud director Addy Osmani, the CLI is “built for humans and agents,” and it already ships with structured JSON output and over 100 bundled agent skills. While setup can be tricky—some users report scope and verification errors—others have wired gws into tools like Claude Code so an AI agent can read, summarize, and act on email from the terminal.
Should Ultra Subscribers Rely on Spark Today?
For people already paying for a Google AI Ultra subscription, Gemini Spark is worth testing as a high-end companion to more traditional task managers and calendar tools. Its strength lies in work that spans services: summarizing your week from Gmail, blocking time in Calendar, and updating Docs or Sheets so everything stays aligned, with far less manual clicking. However, its mixed handling of sensitive or nuanced material means you should treat it as a force multiplier, not an autonomous decision-maker. Keep it focused on structured tasks—like recurring status updates, logistics planning, and inbox triage—and keep human review in the loop for emotional or high-stakes communication. As Google adds features, including financial actions later this summer, Spark’s role may grow from background helper to central nervous system for your digital life, but today it still benefits from close supervision.