Google Antigravity Complete Guide¶

Key Points¶

Agent Supervision UI Design
Agent Manager (Mission Control) sits at the product core, not as an editor add-on
Browser Subagent
Autonomous browser automation agent running in a separate Chrome profile
Governance UI
Terminal Sandboxing, URL Allowlist, Strict Mode — permission control from the settings screen
Agent Skills
SKILL.md-based capability packs auto-selected via Progressive Disclosure

📖 What is Antigravity?¶

Google Antigravity is a Gemini-powered AI-driven integrated development environment. While traditional AI IDEs add AI chat as an auxiliary to the editor, Antigravity places Agent Manager at the product core, integrating management → evidence → verification → review as a single operational unit.

Key Features¶

Manager View: Dashboard for simultaneously supervising multiple workspaces × multiple agents
Browser Subagent: Autonomous browser operation via Gemini 2.5 Pro UI Checkpoint
Artifacts: Auto-structured task lists, implementation plans, screenshots, and recordings
Governance UI: GUI-based permission settings for terminal/browser/workspace
Agent Skills: SKILL.md-based extensible capability packs
VS Code Compatible: Extension and keybinding migration supported
Multi-model: Gemini 3 Pro, Deep Think, Claude Opus 4.5, etc. selectable by use case

Key Specifications (March 2026)¶

Item	Details
Base	VS Code-compatible editor + Agent Manager
Browser	Separate Chrome profile, DOM capture, video recording
Security	Strict Mode (workspace isolation + network disconnection)
Plans	Free preview / Google AI Pro ($20/month)
Target	Personal Gmail accounts (enterprise expansion in progress)

🔀 Which Article Should You Read?¶

graph TD
    Q1{"Have you used<br>Antigravity before?"}
    Q2{"What do you want<br>to learn?"}

    A1["🔍 Overview<br>→ In-Depth Review"]
    A2["🏗️ Design /<br>Security"]
    A3["🚀 Comparison<br>with other tools"]
    A4["🎯 Skills"]

    Q1 -- No --> A1
    Q1 -- Yes --> Q2
    Q2 -- Architecture --> A2
    Q2 -- Comparison --> A3
    Q2 -- Skills --> A4

    style A1 fill:#e8eaf6,stroke:#3f51b5
    style A2 fill:#fff3e0,stroke:#f57c00
    style A3 fill:#e8eaf6,stroke:#3f51b5
    style A4 fill:#e8eaf6,stroke:#3f51b5

Result	Summary
In-Depth Review	Comprehensive feature review
Architecture Deep Dive	Design philosophy, governance UI, security settings
Strengths and Trade-offs	Structural comparison with Cursor and Claude Code
Skills Guide	Agent Skills creation and operations

📋 Goal-Based Guide¶

🔰 Getting the Big Picture¶

Goal	Article
Understand the full feature set	In-Depth Review
Compare with other AI IDEs	Strengths and Trade-offs
Get a high-level overview	Current page (this article)

🏗️ Architecture & Security¶

Goal	Article
Understand design philosophy in depth	Architecture Deep Dive
Review governance UI settings	Architecture > Governance UI
Understand Strict Mode	Architecture > Strict Mode
Learn Browser Subagent authentication	Architecture > Browser Subagent

🎯 Skills & Extensions¶

Goal	Article
Learn skills basics	Skills Guide
Understand skill activation principles	Architecture > Agent Skills

Current page (this article)
Comprehensive hub organizing features, settings, and related articles
In-Depth Review
Comprehensive feature review and evaluation
Strengths and Trade-offs
Structural comparison with Cursor and Claude Code
Skills Guide
Agent Skills creation, placement, and conflict avoidance
Architecture Deep Dive
Design philosophy, governance UI, and security deep dive

📚 Next Steps¶

In-Depth Review →

After grasping the big picture, dive into detailed feature reviews

In-Depth Review
Comprehensive feature review
Strengths and Trade-offs
Structural comparison with Cursor and Claude Code
Architecture Deep Dive
Design philosophy, governance UI, and security settings
Skills Guide
Agent Skills creation and operational best practices