Observed Signal · Jan 14, 2026 · Technical Release · Source: OnlineMarketing.de · Impact: 4/5 · Sentiment: Neutral
Google Tests Agent Skills in Antigravity & Auto Browse
Google advances its Gemini-based agent capabilities by introducing Agent Skills in the Antigravity experimentation platform and by testing an Auto Browse tool for Chrome. Agent Skills are reusable knowledge packages that encode how Gemini should approach specific tasks, described as an open standard to extend agent behavior across projects and contexts. The aim is to standardize workflows, store best practices, templates, and resources as modular skills, reducing dependence on single prompts. Separately, Google is prototyping Auto Browse, enabling Gemini to autonomously open sites, switch tabs, compare sources, and collect information, with code hints indicating Chrome-side integration and initial availability for Gemini Ultra-tier users. Together, these developments illustrate Google's push to move AI agents from passive responders to proactive, browser-enabled executors across Search, Gmail, and related services.
Major platform technical release from Google introducing agentic capabilities and browser automation; high industry relevance
Track Google Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Agent Skills introduced in Google Antigravity to define reusable Gemini capabilities
- Agent Skills described as an open standard to package knowledge into reusable skills
- Auto Browse tested to autonomously browse Chrome, open sites, switch tabs, and compare sources
- Auto Browse planned for Chrome side panel, initially available to Gemini Ultra-tier users
- Gemini aims to serve as a central integration across search, Gmail, browser, and advertising
Connected Companies & Entities
2 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Chrome adds Gemini-based browsing agent for Workspace
Google announced at Google Cloud Next a Gemini-powered browsing agent (branded as an expansion of the "Auto Browse" feature) for Chrome aimed at Google Workspace users. The sidebar agent can read open tabs and assist with tasks such as transferring data from Google Docs to CRMs, cross-tab price comparisons, summarizing candidate portfolios, and extracting product data; users can save repeatable workflows called "Skills." The rollout is initially limited to Workspace customers in the U.S. and follows an earlier January launch for AI‑Pro and AI‑Ultra subscribers. Google emphasized security controls: enterprise inputs will not be used to train its models and every agent action requires human review. Chrome Enterprise Premium will also gain "Shadow‑IT risk detection" to surface approved and unapproved AI/SaaS usage in organizations.
Google Launches Official Agent Skills Repository
Google announced an official Agent Skills repository at Google Cloud Next 2026. The repository (github.com/google/skills) ships with thirteen skills covering product-specific knowledge (AlloyDB, BigQuery, Cloud Run, Cloud SQL, Firebase, the Gemini API, GKE), three Well‑Architected pillars (Security, Reliability, Cost Optimization), and three recipe skills (onboarding, authentication, network observability). Agent Skills is an open format originally developed by Anthropic; it packages condensed, agent-optimized expertise that loads on demand using progressive disclosure to reduce "context bloat" that arises from pulling large documentation into model contexts. The skills are versioned, auditable, composable, installable via a single command, and compatible with Antigravity, Gemini CLI, and third‑party agents. Google says more skills will be added over coming weeks and months.
Google adds Gemini 'auto browse' to Chrome for enterprise
At Google Cloud Next, Google announced it will add Gemini-powered "auto browse" agentic capabilities to Chrome for enterprise Workspace users, allowing the browser to understand open-tab context and automate web tasks such as booking travel, data entry, scheduling, vendor comparisons and summarization. Workflows require a human-in-the-loop to review and confirm actions. The feature will initially roll out to Workspace customers in the U.S. and can be enabled via admin policy; Google says organization prompts will not be used to train its models. Chrome will let users save reusable workflows called "Skills." Separately, Google expanded Chrome Enterprise Premium security features — including "Shadow IT risk detection" and anomalous agent activity monitoring — and announced tighter integrations with Okta and Microsoft Information Protection to reduce session hijacking and enforce security policies.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.