# Table of Contents - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [404 Not Found](#404-not-found) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) - [Google Antigravity Documentation](#google-antigravity-documentation) --- # Google Antigravity Documentation * side\_navigation * [Home](https://antigravity.google/docs/home) \>* Getting Started Getting Started =============== Download -------- Please visit [antigravity.google/download](https://antigravity.google/download) to download Google Antigravity. * **macOS**: macOS versions with Apple security update support. This is typically the current and two previous versions. Min Version 12 (Monterey), X86 is not supported * **Windows**: Windows 10 (64 bit) * **Linux**: glibc >= 2.28, glibcxx >= 3.4.25 (e.g. Ubuntu 20. Debian 10, Fedora 36, RHEL 8) The application will prompt when updates are available: ![Update Available](https://antigravity.google/docs/assets/image/docs/restart-to-update.png) Basic Navigation ---------------- The Agent Manager can be opened from the Editor via the button on the top bar or via keyboard shortcut `Cmd + E`: ![Editor Open Agent Manager](https://antigravity.google/docs/assets/image/docs/editor-open-agent-manager.png) Similarly, from the Agent Manager, the Editor can be opened from any workspace via the “Focus Editor” option in the workspace’s drop down. When focused on a workspace, the Editor can be opened from any of the “Open Editor” buttons, or via the keyboard shortcut `Cmd + E`. ![Agent Manager Open Editor](https://antigravity.google/docs/assets/image/docs/agent-manager-open-editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Home](https://antigravity.google/docs/home) \>* Getting Started Getting Started =============== Download -------- Please visit [antigravity.google/download](https://antigravity.google/download) to download Google Antigravity. * **macOS**: macOS versions with Apple security update support. This is typically the current and two previous versions. Min Version 12 (Monterey), X86 is not supported * **Windows**: Windows 10 (64 bit) * **Linux**: glibc >= 2.28, glibcxx >= 3.4.25 (e.g. Ubuntu 20. Debian 10, Fedora 36, RHEL 8) The application will prompt when updates are available: ![Update Available](https://antigravity.google/docs/assets/image/docs/restart-to-update.png) Basic Navigation ---------------- The Agent Manager can be opened from the Editor via the button on the top bar or via keyboard shortcut `Cmd + E`: ![Editor Open Agent Manager](https://antigravity.google/docs/assets/image/docs/editor-open-agent-manager.png) Similarly, from the Agent Manager, the Editor can be opened from any workspace via the “Focus Editor” option in the workspace’s drop down. When focused on a workspace, the Editor can be opened from any of the “Open Editor” buttons, or via the keyboard shortcut `Cmd + E`. ![Agent Manager Open Editor](https://antigravity.google/docs/assets/image/docs/agent-manager-open-editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Home](https://antigravity.google/docs/home) \>* Getting Started Getting Started =============== Download -------- Please visit [antigravity.google/download](https://antigravity.google/download) to download Google Antigravity. * **macOS**: macOS versions with Apple security update support. This is typically the current and two previous versions. Min Version 12 (Monterey), X86 is not supported * **Windows**: Windows 10 (64 bit) * **Linux**: glibc >= 2.28, glibcxx >= 3.4.25 (e.g. Ubuntu 20. Debian 10, Fedora 36, RHEL 8) The application will prompt when updates are available: ![Update Available](https://antigravity.google/docs/assets/image/docs/restart-to-update.png) Basic Navigation ---------------- The Agent Manager can be opened from the Editor via the button on the top bar or via keyboard shortcut `Cmd + E`: ![Editor Open Agent Manager](https://antigravity.google/docs/assets/image/docs/editor-open-agent-manager.png) Similarly, from the Agent Manager, the Editor can be opened from any workspace via the “Focus Editor” option in the workspace’s drop down. When focused on a workspace, the Editor can be opened from any of the “Open Editor” buttons, or via the keyboard shortcut `Cmd + E`. ![Agent Manager Open Editor](https://antigravity.google/docs/assets/image/docs/agent-manager-open-editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Home](https://antigravity.google/docs/home) \>* Getting Started Getting Started =============== Download -------- Please visit [antigravity.google/download](https://antigravity.google/download) to download Google Antigravity. * **macOS**: macOS versions with Apple security update support. This is typically the current and two previous versions. Min Version 12 (Monterey), X86 is not supported * **Windows**: Windows 10 (64 bit) * **Linux**: glibc >= 2.28, glibcxx >= 3.4.25 (e.g. Ubuntu 20. Debian 10, Fedora 36, RHEL 8) The application will prompt when updates are available: ![Update Available](https://antigravity.google/docs/assets/image/docs/restart-to-update.png) Basic Navigation ---------------- The Agent Manager can be opened from the Editor via the button on the top bar or via keyboard shortcut `Cmd + E`: ![Editor Open Agent Manager](https://antigravity.google/docs/assets/image/docs/editor-open-agent-manager.png) Similarly, from the Agent Manager, the Editor can be opened from any workspace via the “Focus Editor” option in the workspace’s drop down. When focused on a workspace, the Editor can be opened from any of the “Open Editor” buttons, or via the keyboard shortcut `Cmd + E`. ![Agent Manager Open Editor](https://antigravity.google/docs/assets/image/docs/agent-manager-open-editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * FAQ FAQ === Why can I not authenticate into Google Antigravity? --------------------------------------------------- Google Antigravity is currently available for personal Google accounts in approved geographies. Please try using an @gmail.com email address if having challenges with Workspace Google accounts. Why is my age unverified? ------------------------- At the moment, Antigravity is unavailable to under-18 users. If you do meet the minimum age requirement, you may [verify your age](https://myaccount.google.com/age-verification) to continue using Antigravity. What is Google Antigravity’s geographical availability? ------------------------------------------------------- Google Antigravity is available in the following countries and territories. If you're not in one of these countries or territories, you will be unable to use Google Antigravity at this time: **Important**: Please check the country listed on the [Google Terms of Service](https://policies.google.com/terms) page. If this is the wrong country, you may [submit a request](https://policies.google.com/country-association-form) to change your associated region. Americas * American Samoa * Anguilla * Antigua and Barbuda * Argentina * Aruba * The Bahamas * Barbados * Belize * Bermuda * Bolivia * Brazil * British Virgin Islands * Canada * Caribbean Netherlands * Cayman Islands * Chile * Colombia * Costa Rica * Curaçao * Dominica * Dominican Republic * Ecuador * El Salvador * Falkland Islands (Islas Malvinas) * Greenland * Grenada * Guatemala * Guyana * Haiti * Honduras * Jamaica * Mexico * Montserrat * Nicaragua * Panama * Paraguay * Peru * Puerto Rico * Saint Barthélemy * Saint Kitts and Nevis * Saint Lucia * Saint Pierre and Miquelon * Saint Vincent and the Grenadines * South Georgia and the South Sandwich Islands * Suriname * Trinidad and Tobago * Turks and Caicos Islands * United States * Uruguay * U.S. Virgin Islands * Venezuela Europe * Albania * Armenia * Austria * Azerbaijan * Belgium * Bosnia and Herzegovina * Bulgaria * Croatia * Cyprus * Czech Republic * Denmark * Estonia * Faroe Islands * Finland * France * Georgia * Germany * Gibraltar * Greece * Guernsey * Hungary * Iceland * Ireland * Isle of Man * Italy * Jersey * Kosovo * Latvia * Liechtenstein * Lithuania * Luxembourg * Malta * Montenegro * Netherlands * North Macedonia * Norway * Poland * Portugal * Romania * Serbia * Slovakia * Slovenia * Spain * Sweden * Switzerland * Ukrainian territories other than Crimea, the so-called Donetsk People's Republic ("DNR"), and the so-called Luhansk People's Republic ("LNR") * United Kingdom Africa * Algeria * Angola * Benin * Botswana * Burkina Faso * Burundi * Cabo Verde * Cameroon * Central African Republic * Chad * Comoros * Côte d'Ivoire * Democratic Republic of the Congo * Djibouti * Egypt * Equatorial Guinea * Eritrea * Eswatini * Ethiopia * Gabon * The Gambia * Ghana * Guinea * Guinea-Bissau * Kenya * Lesotho * Liberia * Libya * Madagascar * Malawi * Mali * Mauritania * Mauritius * Morocco * Mozambique * Namibia * Niger * Nigeria * Republic of the Congo * Rwanda * Saint Helena, Ascension and Tristan da Cunha * São Tomé and Príncipe * Senegal * Seychelles * Sierra Leone * Somalia * South Africa * South Sudan * Sudan * Tanzania * Togo * Tunisia * Uganda * Western Sahara * Zambia * Zimbabwe Asia * Bahrain * Bangladesh * Bhutan * British Indian Ocean Territory * Brunei * Cambodia * Christmas Island * Cocos (Keeling) Islands * India * Indonesia * Iraq * Israel * Japan * Jordan * Kazakhstan * Kuwait * Kyrgyzstan * Laos * Lebanon * Malaysia * Maldives * Mongolia * Nepal * Oman * Pakistan * Palestine * Philippines * Qatar * Saudi Arabia * Singapore * South Korea * Sri Lanka * Taiwan * Tajikistan * Thailand * Timor-Leste * Türkiye * Turkmenistan * United Arab Emirates * Uzbekistan * Vietnam * Yemen Oceania * Australia * Cook Islands * Fiji * Guam * Heard Island and McDonald Islands * Kiribati * Marshall Islands * Micronesia * Nauru * New Caledonia * New Zealand * Niue * Norfolk Island * Northern Mariana Islands * Palau * Papua New Guinea * Pitcairn Islands * Samoa * Solomon Islands * Tokelau * Tonga * Tuvalu * United States Minor Outlying Islands * Vanuatu * Wallis and Futuna Antarctica * Antarctica Why am I ineligible for a Google One AI plan? --------------------------------------------- The following regions do not currently have access to Google One AI plans: * Antarctica * Brunei * Caribbean Netherlands * Curaçao * Democratic Republic of the Congo * Eswatini * Falkland Islands (Islas Malvinas) * Faroe Islands * Greenland * Guernsey * Heard Island and McDonald Islands * Isle of Man * Jersey * Kosovo * Montenegro * New Caledonia * Pitcairn Islands * Republic of the Congo * Saint Barthélemy * Saint Helena, Ascension and Tristan da Cunha * Saint Kitts and Nevis * Saint Vincent and the Grenadines * São Tomé and Príncipe * South Georgia and the South Sandwich Islands * South Sudan * Sudan * The Gambia * Türkiye * United States Minor Outlying Islands * U.S. Virgin Islands * Wallis and Futuna What is Google Antigravity’s stance on data collection? ------------------------------------------------------- Please refer to the [Terms of Service](https://antigravity.google/terms) . You may opt out of data collection at any point from the Settings panel. How do I sign in with a GCP project? ------------------------------------ This sign-in option is only available to invited teams, and we are not currently taking requests. How do I get support? --------------------- Check out the communities on our [Support page](https://antigravity.google/support) . What are the model rate limits? ------------------------------- Please see more details in the docs on [Plans](https://antigravity.google/docs/plans) . Why can’t I use third party software (e.g. Claude Code, OpenClaw, OpenCode) with my Antigravity login? ------------------------------------------------------------------------------------------------------ Using third party software, tools, or services to access Antigravity is a violation of our [Terms of Service](https://antigravity.google/terms) , and severely degrades the experience for legitimate product users. Such actions may be grounds for suspension or termination of your account. If you would like to use a third party coding agent with Gemini, we recommend using a Vertex or AI Studio API key. Does Google Antigravity currently support worktrees? ---------------------------------------------------- Not at the moment. What happens when my computer goes to sleep? -------------------------------------------- If an agent is running, Antigravity will prevent your computer from sleeping. On this Page --- # Google Antigravity Documentation * side\_navigation * [Home](https://antigravity.google/docs/home) \>* Getting Started Getting Started =============== Download -------- Please visit [antigravity.google/download](https://antigravity.google/download) to download Google Antigravity. * **macOS**: macOS versions with Apple security update support. This is typically the current and two previous versions. Min Version 12 (Monterey), X86 is not supported * **Windows**: Windows 10 (64 bit) * **Linux**: glibc >= 2.28, glibcxx >= 3.4.25 (e.g. Ubuntu 20. Debian 10, Fedora 36, RHEL 8) The application will prompt when updates are available: ![Update Available](https://antigravity.google/docs/assets/image/docs/restart-to-update.png) Basic Navigation ---------------- The Agent Manager can be opened from the Editor via the button on the top bar or via keyboard shortcut `Cmd + E`: ![Editor Open Agent Manager](https://antigravity.google/docs/assets/image/docs/editor-open-agent-manager.png) Similarly, from the Agent Manager, the Editor can be opened from any workspace via the “Focus Editor” option in the workspace’s drop down. When focused on a workspace, the Editor can be opened from any of the “Open Editor” buttons, or via the keyboard shortcut `Cmd + E`. ![Agent Manager Open Editor](https://antigravity.google/docs/assets/image/docs/agent-manager-open-editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * Home Google Antigravity ================== Google Antigravity is an agentic development platform, evolving anyone to build in the agent-first era. Antigravity enables developers to operate at a higher, task-oriented level managing agents across workspaces, while also retaining a familiar AI IDE experience for developers. Antigravity extracts agents into their own surface and provides them the tools needed to autonomously operate across the editor, terminal, and browser emphasizing verification and higher-level communication via tasks and artifacts. This capability enables agents to plan and execute more complex, end-to-end software tasks, elevating all aspects of development, from building features, UI iteration, and fixing bugs to research and generating reports. Main Features automatic\_cluster Agent Manager keyboard\_arrow\_right Agent Manager view for an agent-first experience built around planning mode, the conversation UI, and artifact review. automatic\_cluster AI-powered IDE keyboard\_arrow\_right An AI-powered IDE with all of the AI features that developers have come to rely on such as Agent, Tab, and Command. automatic\_cluster Asynchronous Agents keyboard\_arrow\_right Asynchronous, local agents that can work in parallel on all of your workspaces. automatic\_cluster Multi-window keyboard\_arrow\_right A multi-window product with an Editor, Manager, and Browser. automatic\_cluster Browser Agent keyboard\_arrow\_right Agent that can actuate the browser for you and to accomplish dev tasks like dashboard reads, SCM actions, UI testing, etc. in the Browser. Core Surfaces ------------- automatic\_cluster Agent Manager keyboard\_arrow\_right An orchestration “no code” view to start and view tasks in a minimalist product focused on the conversation and artifacts. automatic\_cluster Editor keyboard\_arrow\_right A fully-functional AI-powered IDE that maps to a single workspace. automatic\_cluster Browser keyboard\_arrow\_right Browser-use agent capabilities to read & actuate on more surfaces beyond just the IDE. Key Terms --------- * **Agent**: The primary AI modality within Antigravity. While the user can work tightly with an Agent within the Editor, they can also have multiple agents working across multiple codebases, orchestrated and monitored through the Agent Manager. * **Tab & Command**: The other AI modalities within Antigravity, specifically within the text editor part of the editor surface. Tab is a more powerful “autocomplete” and Command is an inline instructive modality. From past experience, these do not get nearly as much use as the Agent. * **Artifacts**: We define an artifact as anything that the agent creates to allow it to get its work done or communicate its accomplishments to the human user. These include rich markdown files, diff views, architecture diagrams, images, browser recordings, etc. On this Page --- # Google Antigravity Documentation * side\_navigation * Agent Agent ===== The Agent is the main AI functionality within Google Antigravity. It is a multi-step reasoning system powered by a frontier LLM that can reason over your existing code, use a wide range of tools (including the browser), and communicate with the user through tasks, artifacts, and more. Core Components --------------- * Reasoning model * Tools * Artifacts * Knowledge Customizations -------------- * Agent Modes / Settings * MCP * Rules / Workflows You can spin up multiple Agent conversations, including in parallel. Deleting an Agent conversation via “Right Click > Delete Conversation” in the Agent Manager or by clicking the trash icon in the Editor’s Agent panel. On this Page --- # Google Antigravity Documentation * side\_navigation * [Home](https://antigravity.google/docs/home) \>* Firebase Studio Migration Firebase Studio Migration ========================= Antigravity is Google's next-generation, agent-first platform. It’s designed to be the primary home for high-velocity, autonomous development workflows. Instead of relying on just a cloud-based web editor, Antigravity brings the power of AI right into your local development environment. Why Antigravity? ---------------- Antigravity offers significant enhancements over the web-based Code view in Firebase Studio: * **Local environment control**: Antigravity runs locally on your machine, which means you have full control over your filesystem, versions, and terminal. * **True agentic development**: Move beyond basic code completion. Antigravity provides agentic development workflows that can autonomously format, test, and implement entire tasks across your codebase. * **Seamless Firebase support**: You can still easily deploy your projects to Firebase, communicate with Firebase services via the Firebase CLI, and test your functions locally as you always have. Learn how to navigate Antigravity --------------------------------- To help you settle in, here is where you can find your favorite Firebase Studio features in Antigravity: code Cloud IDE Code view keyboard\_arrow\_right Enjoy the familiar interface of VS Code, but supercharged with AI and running locally on your hardware. auto\_awesome Agentic chat keyboard\_arrow\_right The AI chat you used in Firebase Studio is now natively integrated into your IDE, capable of taking autonomous actions across your local files. rocket\_launch App Hosting deployments keyboard\_arrow\_right You can continue to deploy and manage your App Hosting URLs seamlessly using the Firebase CLI or built-in IDE integrations. important\_devices Browser-based emulators keyboard\_arrow\_right Run the Firebase Local Emulator Suite directly on your machine for faster, offline testing. Migrate your Firebase Studio project to Antigravity --------------------------------------------------- Antigravity is a local, agent-first IDE that brings the power of AI into your local development environment. ### Prerequisites Ensure you have the following installed locally and fully up-to-date: * [Google Antigravity IDE](https://antigravity.google/download) * [Node.js](https://nodejs.org/en) (version 20 or higher) * [Firebase CLI](https://firebase.google.com/docs/cli) (version 15.10.0 or higher) ### Step 1: Export and initialize your app **Option 1: Automated migration** This workflow uses the Antigravity agent to autonomously handle project transformation. 1. In Firebase Studio, click the **Move now** button at the top of your workspace. 2. Follow the export method based on the window that appears: * If you see a **Zip and Download** button, click it. * Otherwise, open the command palette (`Cmd` + `Shift` + `P` on Mac or `Ctrl` + `Shift` + `P` on ChromeOS, Windows, or Linux) and run the **Firebase Studio: Zip & Download** command. 1. Extract the folder locally and open it in Antigravity. 2. In the Agent pane within Antigravity, enter the following prompt. To optimize your workflow and conserve tokens, we recommend selecting the **Gemini Flash** model. It’s designed for speed and efficiency in high-volume transformation tasks like file conversion. The Antigravity agent will then begin project migration, requesting your assistance along the way. Follow the agent’s guidance to complete the migration process. If you encounter any errors, prompt the agent to try again. info If the download window doesn’t appear, check your browser’s address bar for a pop-up blocker icon and ensure pop-ups are allowed. **Option 2: Manual export** If you prefer to manage the migration yourself without using AI tokens, you can use the Firebase CLI to manually export your project. This method is direct and does not require agent interaction. Open your terminal and run the following command, replacing `` with the file path to your extracted project folder or the original `.zip` file: warning The `studio:export` command is currently optimized for Next.js, Flutter, and Angular workspaces. While you can use this command for other workspace types, the migration may not be fully successful. We’re actively working to improve the migration flow. ### Step 2: Preview your app Once you have extracted your project and opened it within Antigravity, you can view your application locally: 1. In Antigravity, navigate to the **Run and Debug** menu located in the left sidebar. 2. Click the play button to start your local development server. 3. Follow the instructions in the terminal to preview your app. lightbulb To refine your app or troubleshoot issues, simply chat with the agent using natural language. If the agent pane is hidden, click the Toggle Agent icon at the top of the window to reopen it. ### Step 3: Publish your app Antigravity uses agent skills to publish your app using Firebase best practices. 1. In the chat panel, enter the following prompt: simply instruct the agent: 1. When prompted to run `firebase deploy`, choose **Yes**. The agent will publish to your existing URL if you’ve previously published to App Hosting. If this is your first time publishing to App Hosting, the agent will walk you through the process. 2. For future updates, simply instruct the agent to `publish my app` in the Antigravity chat panel. Continue your work ------------------ There are several ways you can continue your development in Antigravity. * **Running workflows:** In Antigravity, you can seamlessly execute workflows and continue your work with the model by typing `@workflows ` into the agentic chat panel. * **App Hosting deployments:** You can seamlessly deploy your apps directly through the agent using agent skills, or by using the platform-agnostic Firebase CLI and GitHub. * **Troubleshooting:** If you experience deployment issues, try re-authenticating with the Firebase CLI or verifying your project secrets. Thank you for being part of the Firebase Studio journey. Your prototypes and feedback have directly shaped Google's AI tools, and we can’t wait to see what you build next in Antigravity! Need help? ---------- File any migration bugs in our [GitHub Issues](https://github.com/firebase/firebase-tools/issues) . On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Models Models ====== Reasoning Model --------------- For the core reasoning model, Antigravity offers leading frontier models from the Google Vertex Model Garden: * Gemini 3.1 Pro (high) * Gemini 3.1 Pro (low) * Gemini 3 Flash * Claude Sonnet 4.6 (thinking) * Claude Opus 4.6 (thinking) * GPT-OSS-120b Users can select which reasoning model they want to use within the model selector drop down under the conversation prompt box: ![Model Selector Drop Down](https://antigravity.google/assets/image/docs/model-selector.png) The choice of reasoning model is sticky between user messages within a conversation, so if you change the reasoning model while the Agent is running, it will continue to use the previously selected reasoning model until it has completed its steps for that user turn (or until the user cancels the current execution). Learn more about reasoning model rate limits in [our plans page](https://antigravity.google/docs/plans) . Additional Models ----------------- Antigravity uses a number of other models for various parts of the stack that are not customizable: * **Nano Banana Pro 2**: Used by the generative image tool when the Agent wants to produce a UI mockup, needs images to populate a web page or application, generate system or architecture diagrams, or other generative image tasks. * **Gemini 2.5 Pro UI Checkpoint**: Used by the [browser subagent](https://antigravity.google/docs/browser-subagent) to actuate the browser, such as clicking, scrolling, or filling in input. * **Gemini 2.5 Flash**: Used in the background for checkpointing and context summarization. * **Gemini 2.5 Flash Lite**: Used by the codebase semantic search tool. On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Skills Agent Skills ============ Skills are an [open standard](https://agentskills.io/home) for extending agent capabilities. A skill is a folder containing a `SKILL.md` file with instructions that the agent can follow when working on specific tasks. What are skills? ---------------- Skills are reusable packages of knowledge that extend what the agent can do. Each skill contains: * **Instructions** for how to approach a specific type of task * **Best practices** and conventions to follow * **Optional scripts and resources** the agent can use When you start a conversation, the agent sees a list of available skills with their names and descriptions. If a skill looks relevant to your task, the agent reads the full instructions and follows them. Where skills live ----------------- Antigravity supports two types of skills: | Location | Scope | | --- | --- | | `/.agents/skills//` | Workspace-specific | | `~/.gemini/antigravity/skills//` | Global (all workspaces) | **Workspace skills** are great for project-specific workflows, like your team's deployment process or testing conventions. **Global skills** work across all your projects. Use these for personal utilities or general-purpose tools you want everywhere. Note: Antigravity now defaults to .agents/skills, but still maintains backward support for .agent/skills. Creating a skill ---------------- To create a skill: 1. Create a folder for your skill in one of the skill directories 2. Add a `SKILL.md` file inside that folder Every skill needs a `SKILL.md` file with YAML frontmatter at the top: ### Frontmatter fields | Field | Required | Description | | --- | --- | --- | | `name` | No | A unique identifier for the skill (lowercase, hyphens for spaces). Defaults to the folder name if not provided. | | `description` | Yes | A clear description of what the skill does and when to use it. This is what the agent sees when deciding whether to apply the skill. | Tip: Write your description in third person and include keywords that help the agent recognize when the skill is relevant. For example: "Generates unit tests for Python code using pytest conventions." Skill folder structure ---------------------- While `SKILL.md` is the only required file, you can include additional resources: The agent can read these files when following your skill's instructions. How the agent uses skills ------------------------- Skills follow a **progressive disclosure** pattern: 1. **Discovery**: When a conversation starts, the agent sees a list of available skills with their names and descriptions 2. **Activation**: If a skill looks relevant to your task, the agent reads the full `SKILL.md` content 3. **Execution**: The agent follows the skill's instructions while working on your task You don't need to explicitly tell the agent to use a skill—it decides based on context. However, you can mention a skill by name if you want to ensure it's used. Best practices -------------- ### Keep skills focused Each skill should do one thing well. Instead of a "do everything" skill, create separate skills for distinct tasks. ### Write clear descriptions The description is how the agent decides whether to use your skill. Make it specific about what the skill does and when it's useful. ### Use scripts as black boxes If your skill includes scripts, encourage the agent to run them with `--help` first rather than reading the entire source code. This keeps the agent's context focused on the task. ### Include decision trees For complex skills, add a section that helps the agent choose the right approach based on the situation. Example: A code review skill ---------------------------- Here's a simple skill that helps the agent review code: On this Page --- # Google Antigravity Documentation * side\_navigation * Agent Agent Permissions ================= Antigravity uses a unified permission system to control what actions the Agent can perform on your behalf. Every action is represented as a **permission resource** that can be placed into one of three lists: * **Allow**: The action is auto-approved without prompting. * **Deny**: The action is blocked immediately. * **Ask**: The Agent pauses and asks for your approval before proceeding. Resource String Format ---------------------- Every entry in the Allow, Deny, or Ask list follows the format: Where `action` is one of the supported action types and `target` is a pattern describing what the permission covers. ### Supported Actions | Action | Target Format | Matching | | --- | --- | --- | | `command` | `command(prefix)` or `command(*)` | Matches commands by prefix. `command(git)` matches `git add`, `git commit`, etc. | | `read_file` | `read_file(/path)` | Matches the file or everything under the directory. Paths must be literal and absolute. Globs (`*.go`), regex, and `~` are not supported. | | `write_file` | `write_file(/path)` | Same as `read_file`. Also implicitly covers `read_file` for the same path. | | `read_url` | `read_url(domain)` or `read_url(*)` | Matches the domain and all subdomains. Does not match URL paths. | | `mcp` | `mcp(server/tool)`, `mcp(server/*)`, or `mcp(*)` | Matches by exact server name. `server/*` covers all tools on that server. | ### Examples **Allow list** — these actions run without prompting: **Deny list** — these actions are always blocked: **Ask list** — the Agent pauses and asks before proceeding: On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Rules / Workflows Rules ===== Rules are manually defined constraints for the Agent to follow, at both the local and global levels. Rules allow users to guide the agent to follow behaviors particular to their own use cases and style. To get started with Rules: 1. Open the Customizations panel via the "..." dropdown at the top of the editor's agent panel. 2. Navigate to the Rules panel. 3. Click **\+ Global** to create new Global Rules, or **\+ Workspace** to create new Workspace-specific rules. A Rule itself is simply a Markdown file, where you can input the constraints to guide the Agent to your tasks, stack, and style. Rules files are limited to 12,000 characters each. Global Rules ------------ Global rules live in ~/.gemini/GEMINI.md and are applied across all workspaces. Workspace Rules --------------- Workspace rules live in the .agents/rules folder of your workspace or git root. At the rule level you can define how a rule should be activated: * Manual: The rule is manually activated via at mention in Agent’s input box. * Always On: The rule is always applied. * Model Decision: Based on a natural language description of the rule, the model decides whether to apply the rule. * Glob: Based on the glob pattern you define (e.g., _.js, src/\*\*/_.ts), the rule will be applied to all files that match the pattern. Note: Antigravity now defaults to .agents/rules, but still maintains backward support for .agent/rules. @ Mentions ---------- You can reference other files using @filename in a Rules file. If filename is a relative path, it will be interpreted relative to the location of the Rules file. If filename is an absolute path, it will be resolved as a true absolute path, otherwise it will be resolved relative to the repository. For example, @/path/to/file.md will first attempt to be resolved to /path/to/file.md, and if that file does not exist, it will be resolved to workspace/path/to/file.md. Workflows ========= Workflows enable you to define a series of steps to guide the Agent through a repetitive set of tasks, such as deploying a service or responding to PR comments. These Workflows are saved as markdown files, allowing you to have an easy repeatable way to run key processes. Once saved, Workflows can be invoked in Agent via a slash command with the format /workflow-name. While Rules provide models with guidance by providing persistent, reusable context at the prompt level, Workflows provide a structured sequence of steps or prompts at the trajectory level, guiding the model through a series of interconnected tasks or actions. To create a workflow: 1. Open the Customizations panel via the "..." dropdown at the top of the editor's agent panel. 2. Navigate to the Workflows panel. 3. Click the **\+ Global** button to create a new global workflow that can be accessed across all your workspaces, or click the **\+ Workspace** button to create a workflow specific to your current workspace. To execute a workflow, simply invoke it in Agent using the /workflow-name command. You can call other Workflows from within a workflow! For example, /workflow-1 can include instructions like “Call /workflow-2” and “Call /workflow-3”. Upon invocation, Agent sequentially processes each step defined in the workflow, performing actions or generating responses as specified. Workflows are saved as markdown files and contain a title, a description and a series of steps with specific instructions for Agent to follow. Workflow files are limited to 12,000 characters each. Agent-Generated Workflows ------------------------- You can also ask Agent to generate Workflows for you! This works particularly well after manually working with Agent through a series of steps since it can use the conversation history to create the Workflow. On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Agent Modes / Settings Agent Modes / Settings ====================== Conversation-Level ------------------ When starting a new Agent conversation, users can choose between multiple modes: * Planning: Agent can plan before executing tasks. Use for deep research, complex tasks, or collaborative work. In this mode, the Agent organizes its work in [task groups](https://antigravity.google/docs/task-groups) , produces [Artifacts](https://antigravity.google/docs/artifacts) , and takes other steps to thoroughly research, think through, and plan its work for optimal quality. * Fast: Agent will execute tasks directly. Use for simple tasks that can be completed faster, such as renaming variables, kicking off a few bash commands, or other smaller, localized tasks. This is helpful for when speed is an important factor, and the task is simple enough that there is low worry of worse quality. Overall Settings ---------------- Settings across every Agent conversation can be found in the “Agent” tab of the Settings pane. Some of the major ones include: ### Artifact Review Policy These are the possible options for Artifact Review Policy: * Always Proceed: Agent never asks for review * Request Review: Agent always asks for review When Agent decides to request review from the user for implementation plans, this policy determines what the agent does. When set to “Request Review”, the agent will always terminate after notifying, allowing the user to spend time reviewing the plan and adding comments to augment proposed changes. ![Settings Review Policy Proceed](https://antigravity.google/docs/assets/image/docs/agent/settings-review-policy-proceed.png) If you do not need to manually review Agent’s plan before making changes, set this to “Always Proceed”, in which case every time the agent decides to request review from the user, it will then immediately continue with executing the plan. ![Settings Review Policy Manual](https://antigravity.google/docs/assets/image/docs/agent/settings-review-policy-manual.png) ### Terminal Command Auto Execution For the terminal command generation tool: * Request Review: Never auto-execute terminal commands (except those in a configurable Allow list) * Always Proceed: Agent never asks for review (except those in a configurable Deny list) ### Agent Non-Workspace File Access Allow Agent to view and edit files outside of the current workspace. By default, the Agent only has access to the files in the workspace and in the application’s root folder `~/.antigravity/`, which contains [Artifacts](https://antigravity.google/docs/docs/artifacts) , [Knowledge Items](https://antigravity.google/docs/docs/knowledge) , and other Antigravity-specific data. Use with caution, as this could expose local secret or sensitive data to the Agent. On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Sandboxing Sandboxing Terminal Commands ============================ Sandboxing provides kernel-level isolation for terminal commands executed by the Agent. When enabled, commands run in a restricted environment with limited file system and network access, protecting your system from unintended modifications. Sandboxing is currently disabled by default, but this may change in future releases. It is supported on macOS and Linux. On macOS, it leverages Seatbelt (`sandbox-exec`), Apple's kernel-level sandboxing mechanism. On Linux, it uses `nsjail` for process isolation. Enabling Sandboxing ------------------- You can enable or disable sandboxing in Antigravity User Settings. Toggle "Enable Terminal Sandboxing" to turn sandboxing on or off. When enabled, you can also control network access separately using the "Sandbox Allow Network" toggle. ![Sandbox settings toggles](https://antigravity.google/assets/image/docs/sandbox-settings-toggle.png) Restrictions ------------ When sandboxing is enabled, the Agent's terminal commands are subject to the following restrictions: * **File System**: Commands can only write to your designated workspace directory and essential system locations. This prevents the Agent from accidentally deleting or modifying files outside your project. ![File system operation blocked by sandbox](https://antigravity.google/assets/image/docs/sandbox-filesystem-denied.png) * **Network Access**: Network connectivity can be independently controlled. Use the "Sandbox Network Access" toggle in Antigravity User Settings to allow or deny network access while maintaining file system restrictions. Here's an example of a command being blocked due to network restrictions: ![Sandbox network denial example](https://antigravity.google/assets/image/docs/sandbox-network-denied.png) Handling Sandbox Violations --------------------------- If a command fails due to sandbox restrictions, you'll see a message indicating the failure may be sandbox-related. To resolve this, you can: * **Disable Sandbox Permanently**: Turn off sandboxing entirely in Antigravity User Settings. * **Bypass for a Single Command**: When using "Request Review" mode, you can choose to run an individual command with or without sandbox restrictions. In "Request Review" mode, you'll see a "Bypass Sandbox" option when prompted to run a command: ![Bypass Sandbox option in Request Review mode](https://antigravity.google/assets/image/docs/sandbox-bypass-option.png) Interaction with Strict Mode ---------------------------- When strict mode is enabled, sandboxing is automatically activated with network access denied. This ensures maximum protection when operating in a strict environment. ![Sandbox settings in strict mode](https://antigravity.google/assets/image/docs/sandbox-secure-mode-settings.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Artifacts](https://antigravity.google/docs/artifacts) \>* Task List Task List ========= A task list is an artifact that the agent uses to approach complex tasks and monitor progress on various action items. You can find a live snapshot of what the agent is doing in this artifact, which is constructed as a markdown list of items related to research, implementation, verification, and more. This type of artifact is generally used by the agent to keep on track with the user’s overarching goal; typically, you do not need to directly interact with this artifact. ![Task List](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-task.png) On this Page --- # 404 Not Found Error: Not Found ================ The requested URL `/docs/` was not found on this server. -------------------------------------------------------- --- # Google Antigravity Documentation * side\_navigation * Artifacts Artifacts ========= We define an Artifact as anything that the agent creates to allow it to get its work done or communicate its work and thinking to the human user. These include rich markdown files, diff views, architecture diagrams, images, browser recordings, code diffs, etc. As Agents become more autonomous and can run for longer and longer periods, Artifacts allow for the Agent to asynchronously communicate its work to the user, as opposed to requiring the user to carefully monitor every Agent step synchronously. Artifacts are produced while the Agent is in [Planning mode](https://antigravity.google/docs/agent-modes-settings) , and appear in both the [Agent Manager](https://antigravity.google/docs/agent-manager) and [Editor](https://antigravity.google/docs/editor) views, though the former is optimized for displaying, organizing, and managing Artifacts. Feedback is another key concept with Artifacts. Depending on the user settings, the Agent may ask for review on intermediate Artifacts to receive confirmation that it has made progress in its thinking or implementation that aligns with the user’s intent and goal. The user is able to provide feedback on the Artifact to provide guidance to steer the Agent in the proper direction. The UI/UX of feedback differs from Artifact type to Artifact type. On this Page --- # Google Antigravity Documentation * side\_navigation * Tools \>* MCP Antigravity Editor: MCP Integration =================================== Antigravity supports the Model Context Protocol (MCP), a standard that allows the editor to securely connect to your local tools, databases, and external services. This integration provides the AI with real-time context beyond just the files open in your editor. What is MCP? ------------ MCP acts as a bridge between Antigravity and your broader development environment. Instead of manually pasting context (like database schemas or logs) into the editor, MCP allows Antigravity to fetch this information directly when needed. Core Features ------------- ### 1\. Context Resources The AI can read data from connected MCP servers to inform its suggestions. **Example:** When writing a SQL query, Antigravity can inspect your live Neon or Supabase schema to suggest correct table and column names. **Example:** When debugging, the editor can pull in recent build logs from Netlify or Heroku. ### 2\. Custom Tools MCP enables Antigravity to execute specific, safe actions defined by your connected servers. **Example:** "Create a Linear issue for this TODO." **Example:** "Search Notion or GitHub for authentication patterns." How to Connect -------------- Connections are managed directly through the built-in MCP Store. 1. **Access the Store:** Open the MCP Store panel within the "..." dropdown at the top of the editor's side panel. 2. **Browse & Install:** Select any of the supported servers from the list and click Install. 3. **Authenticate:** Follow the on-screen prompts to securely link your accounts (where applicable). Once installed, resources and tools from the server are automatically available to the editor. Connecting Custom MCP Servers ----------------------------- To connect to a custom MCP server: 1. Open the MCP store via the "..." dropdown at the top of the editor's agent panel. 2. Click on "Manage MCP Servers" 3. Click on "View raw config" 4. Modify the mcp\_config.json with your custom MCP server configuration. The configuration file is located at `~/.gemini/antigravity/mcp_config.json`. ### Configuration Structure The configuration file has a single `mcpServers` object where you define each server you want to connect to. ### Configuration Properties Each server entry supports the following properties: **Transport (one required):** * **`command`** (string): Path to the executable for stdio transport. * **`serverUrl`** (string): URL for remote servers for Streamable HTTP transport. **Optional:** * **`args`** (string\[\]): Command-line arguments for stdio transport. * **`env`** (object): Environment variables for the stdio server process. * **`cwd`** (string): Working directory for stdio servers. * **`headers`** (object): Custom HTTP headers for remote servers. * **`authProviderType`** (string): Authentication provider. Supports `"google_credentials"` for ADC. * **`oauth`** (object): OAuth client credentials (`clientId`, `clientSecret`). * **`disabled`** (boolean): Temporarily disable a server without removing its configuration. * **`disabledTools`** (string\[\]): Tool names to not provide to the model. Authentication -------------- ### Google Credentials Set `authProviderType` to `"google_credentials"` to use Google Application Default Credentials (ADC). This requires Application Default Credentials to be configured. To set them up, run: ### OAuth Antigravity can automatically handle OAuth for servers that support dynamic client registration (DCR). For these servers, no additional configuration is needed: If the server does not support dynamic client registration, you can provide your client credentials manually: If you provided client credentials manually, ensure the following is registered as a redirect URI in your OAuth provider: When connecting to an OAuth-enabled server: 1. Open **Agent Settings** with `Cmd+,` (Mac) or `Ctrl+,` (Windows/Linux). 2. Navigate to the **Customizations** tab and click the **Authenticate** button next to the server. ![Click Authenticate](https://antigravity.google/docs/assets/image/docs/tools/mcp-oauth-authenticate.png) 1. Complete authentication in your browser and copy the authorization code. ![Copy authorization code](https://antigravity.google/docs/assets/image/docs/tools/mcp-oauth-copy-code.png) 1. Paste the code back into the settings panel and click **Submit**. ![Paste auth code](https://antigravity.google/docs/assets/image/docs/tools/mcp-oauth-paste-code.png) Once authenticated, the server will reconnect automatically. ![Authenticated server](https://antigravity.google/docs/assets/image/docs/tools/mcp-oauth-authenticated.png) Access tokens are stored in `~/.gemini/antigravity/mcp_oauth_tokens.json`. Expired tokens are refreshed automatically, and invalid tokens are removed. ### Custom Headers For servers that require custom HTTP headers (e.g. API keys or bearer tokens), add them to the `headers` object. For example: Supported Servers ----------------- The MCP Store currently features integrations for: * Airweave * Arize * AlloyDB for PostgreSQL * Atlassian * BigQuery * Chrome DevTools * ClickHouse * Cloud SQL for PostgreSQL * Cloud SQL for MySQL * Cloud SQL for SQL Server * Dart * Dataplex * Figma Dev Mode MCP * Firebase * GitHub * Harness * Heroku * Linear * Locofy * Looker * MCP Toolbox for Databases * MongoDB * Neon * Netlify * Notion * PayPal * Perplexity Ask * Pinecone * Postman * Prisma * Redis * Sequential Thinking * SonarQube * Spanner * Stripe * Supabase On this Page --- # Google Antigravity Documentation * side\_navigation * [Artifacts](https://antigravity.google/docs/artifacts) \>* Walkthrough Walkthrough =========== Agent creates walkthrough artifacts when it has completed task implementation; this type of artifact includes a concise summary of the changes that have been made to remind the user of what has happened in the active conversation. This is a great way to get up to speed with the state of your codebase after Agent has made its changes in case you were not strictly following it the whole time. ![Walkthrough](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-walkthrough.png) For browser tasks, walkthroughs often contain screenshots and screen recordings of what Agent has built or created in the browser! ![Walkthrough with Image](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-walkthrough-image.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Artifacts](https://antigravity.google/docs/artifacts) \>* Implementation Plan Implementation Plan =================== Agent utilizes the implementation plan artifact to architect changes within your codebase to accomplish a task. These plans contain technical details on what revisions are necessary and are meant to be reviewed by the user. Below is an example plan generated by the agent. ![Artifact Implementation Plan](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-implementation-plan.png) Unless you have you artifact review policy set to “Always Proceed” \[link to docs on this setting\], Agent will typically request your review on the implementation plan before making the changes needed to complete your task. You can click either the in-conversation or artifact header “Proceed” button to instantly continue with Agent’s plan. ![Artifact Implementation Plan Proceed](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-implementation-plan-proceed.png) Oftentimes, Agent will create a plan that is slightly different from what you exactly want. Antigravity supports commenting on these artifacts so you can provide feedback to Agent for any reason, whether it be to decrease scope of changes, use a different tech stack, or correct any Agent discrepancies. ![Artifact Implementation Plan Comments](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-implementation-plan-comments.png) Once you have left comments on the implementation plan, you can still use the “Proceed” to continue with Agent’s plan; alternatively, you can also toggle the “Review” button in the artifact header, where you can examine all comments and leave a message as feedback instead of directly proceeding, if needed. ![Artifact Implementation Plan Submit Comments](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-implementation-plan-submit-comments.png) Once you have proceeded or left a review, Agent will continue its work, either iterating on the implementation plan and re-requesting your review or beginning with its work! ![Artifact Implementation Plan Proceeded](https://antigravity.google/docs/assets/image/docs/artifacts/artifact-implementation-plan-proceeded.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Task Groups Task Groups =========== When Agent is in planning mode, large and complex tasks are handled with Task Groups, which break down these problems into smaller, more approachable units of work. Oftentimes, Agent will work on multiple parts of the greater task at the same time, and task sections are how these changes are presented to the user. Here is an example task group. ![Task Group](https://antigravity.google/docs/assets/image/docs/agent/task-group.png) The top component of the task group specifies the overarching goal of this task as well as summarizes the changes made within this unit of work. There is also a section of edited files for quick user audit of changes: click on the file pill and you will view the current state of the changed files. ![Task Group Clicked Pill](https://antigravity.google/docs/assets/image/docs/agent/task-group-clicked-pill.png) Within a task group, Agent identifies subtasks that help modularize necessary changes, and all work done by the Agent is viewable within these progress update sections. By default, the details in each subtask are not directly exposed to the user, but if you are interested, there is a toggle that will expand on the exact steps that Agent made. ![Task Group Expanded](https://antigravity.google/docs/assets/image/docs/agent/task-group-expanded.png) Sometimes, there are pending steps, such as browser setup or terminal commands requiring approval, that are created inside these progress updates. In this case, instead of expanding all of the updates, Agent provides a special section at the end of the task group where you can review these pending steps accordingly. ![Task Group Pending](https://antigravity.google/docs/assets/image/docs/agent/task-group-pending.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Artifacts](https://antigravity.google/docs/artifacts) \>* Knowledge Knowledge ========= Knowledge Items are Antigravity's persistent memory system that automatically captures and organizes important insights, patterns, and solutions from your coding sessions. They help you build upon previous work across conversations. What is a Knowledge Item? ------------------------- A Knowledge Item is a collection of related information on a specific topic. Each Knowledge Item contains a title and summary describing what it covers, and a collection of artifacts providing information on the topic. Possible examples of artifacts include automatically generated documentation, code examples, or persistent memories of user instructions. How are Knowledge Items Generated? ---------------------------------- As you interact with the agent, Antigravity automatically analyzes and extracts information from your conversation and uses that information to create new KIs or update existing KIs. Viewing Knowledge Items ----------------------- You can view your Knowledge Items in the Antigravity **Agent Manager**. ![Knowledge View](https://antigravity.google/docs/assets/image/docs/artifacts/knowledge-view.png) How are Knowledge Items used by the Agent? ------------------------------------------ The summaries of all your Knowledge Items are available to the agent, which uses them to inform its responses. When the agent identifies a Knowledge Item that is relevant to the conversation, it will automatically study the artifacts in that Knowledge Item and use the applicable information. On this Page --- # Google Antigravity Documentation * side\_navigation * [Artifacts](https://antigravity.google/docs/artifacts) \>* Screenshots Screenshots =========== The browser subagent can take screenshots of open pages or elements on pages when it would like your review of the state of the page. This is surfaced as a tool to the agent, and you can also prompt the agent to take a screenshot of a page. ![Browser Screenshot Capture Tool](https://antigravity.google/docs/assets/image/docs/artifacts/browser-screenshot-capture.png) All screenshots are saved as image artifacts and can be commented on to give feedback to the agent. ![Browser Screenshot Artifact](https://antigravity.google/docs/assets/image/docs/artifacts/browser-screenshot-artifact.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Editor](https://antigravity.google/docs/editor) \>* Tab Antigravity Editor: Tab & Navigation ==================================== This guide covers the core navigation and completion tools: **Supercomplete**, **Tab-to-Jump**, and **Tab-to-Import**. Supercomplete ------------- Supercomplete provides code suggestions in a region near your current cursor position. ![Supercomplete](https://antigravity.google/docs/assets/image/docs/editor/supercomplete.png) ### How it Works * **File-Wide Suggestions**: Suggestions can modify code throughout the document, handling tasks like changing variable names or updating separate function definitions simultaneously. * **Accepting**: Press `Tab` to accept the changes. Tab-to-Jump ----------- Tab-to-Jump is a fluid navigation tool that suggests the next logical place in your document to move your cursor to. ![Tab-to-Jump](https://antigravity.google/docs/assets/image/docs/editor/tab_to_jump.png) ### How it Works * A "Tab to jump" icon will appear offering to move your cursor to where your next logical edit will be. Pressing `Tab` instantly moves your cursor to that location. * **Accepting**: Press `Tab` to accept the jump. Tab-to-Import ------------- Tab-to-Import handles missing dependencies without breaking your flow. ![Tab-to-Import](https://antigravity.google/docs/assets/image/docs/editor/tab_to_import.png) ### How it Works * **Detection**: If you type a class or function that isn't imported, Antigravity suggests the import. * **Action**: Press `Tab` to complete the word and instantly add the import statement to the top of the file. Settings -------- In your settings, you can customize the behavior of these features: * **Enable/Disable Features**: You can individually turn off Autocomplete, Tab-to-Jump, Supercomplete, or Tab-to-Import. * **Tab Speed**: Controls the responsiveness of suggestions. * `Slow`: Waits for more context before suggesting. * `Default`: Offers a balanced pace. * `Fast`: Provides rapid-fire suggestions. * **Highlight Inserted Text**: When enabled, text inserted via Tab is highlighted to track changes easily. * **Clipboard Context**: When enabled, Antigravity uses the contents of your clipboard to improve completion accuracy. * **Allow Gitignored Files**: Enables Tab features (suggestions and jumping) within files listed in your `.gitignore` file. Tab will only ignore gitignored files if git is installed. On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Browser Subagent Browser Subagent ================ When the agent wants to interact with the browser, it invokes a browser subagent to handle the task at hand. The browser subagent runs a model specialized to operate on the pages that are open within the Antigravity-managed browser, which is different from the model you selected for the main agent. This subagent has access to a variety of tools that are necessary to control your browser, including clicking, scrolling, typing, reading console logs, and more. It can also read your open pages through DOM capture, screenshots, or markdown parsing, as well as taking videos. While the agent is controlling a page, it will show an overlay on the page with a blue border and a small panel with short descriptions of the actions being taken. When this is shown, you will not be allowed to interact with the page to ensure it doesn’t get confused by your actions. The browser subagent can act on tabs that are not focused, so you are free to open other tabs and use them uninterrupted as it works. On this Page --- # Google Antigravity Documentation * side\_navigation * [Artifacts](https://antigravity.google/docs/artifacts) \>* Browser Recordings Browser Recordings ================== Every time the browser subagent actuates on the Browser, it may choose to generate a recording of the agent’s actions for your review. You can view this playback, if it is available, at the bottom of the Browser step UI. ![Browser Recording Capture Tool](https://antigravity.google/docs/assets/image/docs/artifacts/browser-recording-capture.png) All browser recordings are also saved as a recording artifact for your review. This view loops through the browser agent’s actions. ![Browser Recording Artifact](https://antigravity.google/docs/assets/image/docs/artifacts/browser-recording-artifact.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Conversation View](https://antigravity.google/docs/conversation-view) \>* Files Files ===== As you open up [file panes](https://antigravity.google/docs/panes) within the manager, you can also leave comments for the agent to highlight specific points. ![File commenting manager](https://antigravity.google/docs/assets/image/docs/file_commenting_manager.png) On this Page --- # Google Antigravity Documentation * side\_navigation * Browser Browser ======= Antigravity has the ability to open, read, and control a Chrome browser, allowing you to test development websites, read internet data sources, and automate various different browser tasks. Using a [subagent](https://antigravity.google/docs/browser-subagent) , Antigravity can operate on the browser as it sees fit, as well as [recording its actions](https://antigravity.google/docs/browser-recordings) and presenting select screenshots and videos as [artifacts](https://antigravity.google/docs/screenshots) . To isolate the Antigravity agent from your normal browsing, it runs in a separate browser profile. This means that it will show up as a separate application in your dock and nothing will be signed in. You can read more about this in the [Separate Chrome Profile](https://antigravity.google/docs/separate-chrome-profile) section. To disable the use of all browser tools, you may disable the browser tools setting in the “Browser” section of your settings. ![Enable Browser Tools](https://antigravity.google/assets/image/docs/enable_browser_tools.png) Antigravity detects and uses your existing Chrome application. If you don’t already have Chrome, you must download it here. If Antigravity is unable to detect your chrome installation, you may have to specify the path to it in the following setting: ![Chrome Binary Path](https://antigravity.google/assets/image/docs/chrome_binary_path.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Browser](https://antigravity.google/docs/browser) \>* Allowlist / Denylist Allowlist / Denylist ==================== The browser uses a two-layer security system to control which URLs can be accessed: * **Denylist** - Deny dangerous/malicious URLs * **Allowlist** - Explicitly allow trusted URLs How It Works ------------ ### Denylist The denylist is maintained and enforced using the Google Superroots’s BadUrlsChecker service (See documentation). When the browser attempts to navigate to a URL, the hostname is checked against the server-side denylist via RPC. **NOTE:** If the server is unavailable, access is denied by default. ### Allowlist The allowlist is a local text file that you can edit to explicitly trust specific URLs. ![Allowlist](https://antigravity.google/docs/assets/image/docs/browser-allowlist.png) The allowlist is initialized with just localhost, and can be edited at anytime. When the browser attempts to navigate to a non-allowlisted URL, it will prompt you with an “always allow” button, which if clicked will add the URL to the allowlist and enable the browser to open and interact with the web page. An example situation is shown below: ![Always Allow](https://antigravity.google/docs/assets/image/docs/always-allow-url.png) You can also add/remove URLS from the allowlist manually. However, the denylist always takes precedence: you cannot allowlist a URL that appears on the denylist. On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent](https://antigravity.google/docs/agent) \>* Strict Mode Strict Mode =========== Strict mode provides enhanced security controls for the Agent, allowing you to restrict its access to external resources and sensitive operations. When strict mode is enabled, several security measures are enforced to protect your environment. Features -------- ### Browser URL Allowlist/Denylist In strict mode, the Agent's ability to interact with external websites is governed by the browser's Allowlist and Denylist. This applies to: * **External Markdown Images**: The Agent will only render images from URLs that are allowed. * **Read URL Tool**: The Read URL tool will only auto-execute for allowed URLs. ### Terminal, Browser, and Artifact Review Policies Strict mode enforces the following behavior for terminal, browser, and artifact interactions: * **Terminal Auto Execution**: Set to "Request Review". The Agent will always prompt for permission before executing any terminal command. The terminal allowlist is ignored when strict mode is enabled. * **Browser Javascript Execution**: Set to "Request Review". The Agent will always prompt for permission before executing Javascript in the browser. * **Artifact Review**: Set to "Request Review". The Agent will always prompt for confirmation before acting on plans laid out in artifacts. ### File System Access Strict mode restricts the Agent's access to the file system to ensure it only interacts with authorized files: * **Respect .gitignore**: The Agent will respect `.gitignore` rules, preventing it from accessing ignored files. * **Workspace Isolation**: Access to files outside the workspace is disabled. The Agent can only view and edit files within the designated workspace. On this Page --- # Google Antigravity Documentation * side\_navigation * [Editor](https://antigravity.google/docs/editor) \>* Command Antigravity Editor: Command =========================== The **Command** feature brings the power of natural language directly into your workflow, allowing you to request specific inline completions or terminal commands on the fly. How it Works ------------ 1. **Trigger**: Press `Command + I` (Mac) or `Ctrl + I` (Windows/Linux). 2. **Prompt**: A text input box will appear at your current cursor position. 3. **Instruction**: Type your request in natural language (e.g., "Create a function to sort this list" or "Add error handling to this block"). 4. **Execution**: Antigravity generates the code or command directly inline for you to review and accept. Where to Use It --------------- ### In the Editor ![Command in Editor](https://antigravity.google/docs/assets/image/docs/editor/command_editor.png) Use Command to generate boilerplate code, refactor complex functions, or write documentation without breaking your coding flow. * _Example_: "Create a React component for a login form." ### In the Terminal ![Command in Terminal](https://antigravity.google/docs/assets/image/docs/editor/command_terminal.png) Use Command within the integrated Antigravity terminal to generate complex shell commands without needing to memorize syntax. * _Example_: "Find all processes listening on port 3000 and kill them." On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* Workspaces Workspaces ========== In the Agent Manager, you can work across multiple workspaces simultaneously. In order to open a new workspace, just select the button in the left sidebar and select a starting folder. At any point, you can switch between conversations across workspaces through the left sidebar. ![Switch Workspaces](https://antigravity.google/assets/image/docs/workspaces/switch_workspace.png) To start a new conversation within a workspace, either select the desired workspace from the Start Conversation tab or hit the Plus button next to the workspace name in the sidebar. ![Start Conversation Within Workspace](https://antigravity.google/assets/image/docs/workspaces/start_within_workspace.png) On this Page --- # Google Antigravity Documentation * side\_navigation * Agent Manager Agent Manager ============= We've built out the Agent Manager, to provide a higher level view into the work Antigravity agents are doing under your guidance. Here, you can work across multiple workspaces, oversee dozens of agents simultaneously, and interact with your codebase primarily through the agent, rather than through writing code directly. As agents and models continue to get better, we believe that this birds-eye view will be the primary entry point to all of your work. We expect to move it to be central to the Antigravity experience soon. At any point, you can toggle between the Agent Manager and the editor by hitting CMD+E (Mac) or CTRL+E (Windows), or through the Open Editor & Open Agent Manager buttons at the top right of the menu bar. You can also manage your editor windows through the manager, either hiding, focusing, or closing them. ![Agent Manager Editor](https://antigravity.google/docs/assets/image/docs/agent_manager_editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Workspaces](https://antigravity.google/docs/workspaces) \>* Inbox Inbox ===== The inbox is your one stop shop to track all of your conversations in one place. From the inbox you can see if any of your conversations are awaiting your approval to run terminal commands, use the browser, or build out an implementation plan. ![Inbox Overview](https://antigravity.google/assets/image/docs/inbox/basic_inbox.png) You can use the search bar and the pending switch to search for conversations by folder or by title to make sure your inbox is always focused on what is most relevant to you. Selecting a conversation from the inbox will jump directly to the conversation, where you can continue where you left off. ![Continue Conversation](https://antigravity.google/assets/image/docs/inbox/continue_conversation.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* Conversation View Conversation View ================= The [Agent Panel](https://antigravity.google/docs/agent-side-panel) takes center stage in the agent manager. As the agent makes progress, you'll be able to follow along with what it's doing. To toggle off this follow-along mode, simply hit the `Following` button at the top right of the conversation. ![Follow Along Manager](https://antigravity.google/docs/assets/image/docs/follow_along_manager.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Conversation View](https://antigravity.google/docs/conversation-view) \>* Review Changes + Source Control Review Changes + Source Control =============================== Just as in the [editor](https://antigravity.google/docs/review-changes-editor) , you can easily review the work you and your agent have collaborated on from within the manager. Once you enter a conversation, you can toggle the Review Changes pane through the button at the top right to open up a pane where you can scroll through and comment on any file diffs made as a part of the conversation. ![Review Changes Manager](https://antigravity.google/docs/assets/image/docs/review_changes_manager.png) You can similarly toggle to the Source Control tab within the Review Changes pane to see changed files, stage or unstage them, and commit them upstream. ![Manager Source Control](https://antigravity.google/docs/assets/image/docs/manager_source_control.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Workspaces](https://antigravity.google/docs/workspaces) \>* Playground Playground ========== Playgrounds are independent workspaces that allow you to start a conversation and explore ideas instantly, without the overhead of setting up a new workspace. Creating a Playground --------------------- From the Start Conversation page, you can quickly send a message to a new playground by clicking the Use Playground button below the input box. ![Playground Button](https://antigravity.google/assets/image/docs/playground/playground_button.png) From here, you can send a message as you would normally. ![Playground Send Message](https://antigravity.google/assets/image/docs/playground/playground_send.png) Persisting Your Work -------------------- If you want to keep the work you've done in a playground, you can move its contents into a dedicated workspace in a folder of your choosing. This action preserves your conversation history and any files created during your session, allowing you to continue exploring with multiple conversations. ![Playground Persist](https://antigravity.google/assets/image/docs/playground/playground_persist.png) Clicking on the move button in the top bar will pop up the following modal: ![Playground Move Modal](https://antigravity.google/assets/image/docs/playground/playground_move.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Conversation View](https://antigravity.google/docs/conversation-view) \>* Panes Panes ===== You can open files, [artifacts](https://antigravity.google/docs/artifacts) , knowledge items, and other content directly within the manager in panes that persist per-conversation. In order to open up a pane, simply open up the quick picker (by hitting CMD+P on Mac or CTRL+P on Windows/Linux) and select a resource. You can also hit the "+" from within a conversation's header. ![Manager New Tab](https://antigravity.google/docs/assets/image/docs/agent-manager/manager_new_tab.png)![Manager File Picker](https://antigravity.google/docs/assets/image/docs/agent-manager/manager_file_picker.png) These panes are resizable, splittable, and drag-and-droppable. You should configure them around as makes sense for your workflow. ![Manager Split Pane](https://antigravity.google/docs/assets/image/docs/agent-manager/manager_split_pane.png) If you use CMD+Click or CMD+Enter (Mac) or CTRL+Click / CTRL+Enter (non-Mac), the contents will open in a new pane, rather than replacing the currently open pane. On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Conversation View](https://antigravity.google/docs/conversation-view) \>* Browser Subagent View Browser Subagent View ===================== Overview -------- The Manager has a dedicated side panel that allows you to expand and inspect the Agent’s work for a task. ![Browser Subagent View](https://antigravity.google/docs/assets/image/docs/browser-subagent-view.png) In the regular manager conversation view (left half of the image), the browser subagent’s work is hidden. Clicking the expand button (shown in red box) will bring up the subagent view (right half of the image). Updates to the Agent’s work will be streamed into this view, so you can follow along and interact with steps as required (e.g. confirm/deny actions). What’s in the side panel ------------------------ * All subagent actions (clicking, scrolling, navigating, etc.) * Visual feedback showing exactly where clicks happened * Screenshots captured at each step Visual Inspection Feature ------------------------- Tool calls that produce actions in the browser, like clicks, include a button (shown in blue box) which opens a screenshot of the browser at that exact moment and a red dot showing what interaction the agent has done in the browser. On this Page --- # Google Antigravity Documentation * side\_navigation * [Browser](https://antigravity.google/docs/browser) \>* Separate Chrome Profile Separate Chrome Profile ======================= To isolate the browser from your general browsing, it operates on a [separate Chrome profile](https://support.google.com/chrome/answer/2364824) . Since Chrome profiles are isolated, this will not share any of the cookies or sign-in information from your normal browsing profile. However, all sign-ins will be persisted such that anytime you open the browser in the future, all your accounts will still be there. If you had your normal Chrome open while launching this profile, it will show up as a separate dock icon and be considered a separate application. If Chrome was not open beforehand, this application will look the same as your default profile. To return to the default profile, you must quit the application and relaunch Chrome. If you would like to change the location where your browser profile will be created, you can modify the following setting in the browser section. ![Browser Profile](https://antigravity.google/assets/image/docs/browser/browser-profile.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Conversation View](https://antigravity.google/docs/conversation-view) \>* Changes Sidebar Changes Sidebar =============== Similar to the toolbar at the bottom of the editor's [Agent Panel](https://antigravity.google/docs/docs/agent-side-panel) , the Changes Sidebar in the manager offers a quick way to see what [artifacts](https://antigravity.google/docs/docs/artifacts) the agent has created and what files it has modified within a conversation. Clicking on any of the listed resources will open its contents within a pane. The icons on each resource indicate whether there are new changes to a resource since your last review. ![Aux Sidebar](https://antigravity.google/docs/assets/image/docs/aux-sidebar.png) On this Page --- # Google Antigravity Documentation * side\_navigation * Settings Settings ======== You can configure your Antigravity settings across Agent, Browser, Editor, and more via: * Keyboard shortcut in any surface: `Cmd + ,` * From the Settings tab or gear icon in the Agent Manager * From "Settings > Open Antigravity User Settings" in the Editor Data Collection Settings ------------------------ The "Enable Telemetry" setting can be found in the Settings panel under the "Account" section. When toggled on, Antigravity collects interactions for use in evaluating, developing, and improving Antigravity and models that support Antigravity. On this Page --- # Google Antigravity Documentation * side\_navigation * Plans Plans ===== At this moment, Google Antigravity is available with [terms](https://antigravity.google/terms) to individual accounts derived from Google's terms of service, and in preview (pre-general availability) to teams derived from Section 5 of the General Service Terms in Google Cloud's enterprise terms of service. Rate limits and model availability differs based on usage of [Google AI](https://one.google.com/about/google-ai-plans/) plans. Baseline Quota -------------- All plans receive a baseline of: * Use of Gemini 3 Pro, Gemini 3 Flash, and other offered Vertex Model Garden models as the core agent model * Unlimited Tab completions * Unlimited Command requests * Access to all product features, such as the Agent Manager and Browser integration Users on Google AI Ultra receive: * The highest, most generous quota, refreshed every five hours * Highest weekly rate limits Users on Google AI Pro receive: * High, generous quota, refreshed every five hours until weekly limit reached * Higher weekly rate limit Users not on AI Pro and Ultra plans receive: * Meaningful quota, refreshed weekly * Weekly rate limit The baseline rate limits are primarily determined to the degree we have capacity, and exist to prevent abuse. Under the hood, the rate limits are correlated with the amount of work done by the agent, which can differ from prompt to prompt. Thus, you may get many more prompts if your tasks are more straightforward and the agent can complete the work quickly, and the opposite is also true. Usage limits for this service are subject to modification. These adjustments may be necessary to manage system capacity and maintain service stability. Overages -------- Users on Google AI Pro or Ultra plans also have the ability to utilize their plan-included AI credits for additional overage usage above the baseline provided quota, and can [purchase additional AI credits](http://one.google.com/ai/credits) if desired. AI credits are consumed at Vertex API pricing. Usage of credits once the baseline quota is exhausted for any particular model is controlled by the "AI Credit Overages" user setting, which can be set to the following: * Never: Never use AI credits automatically, wait until the baseline quota refreshes before using this model further * Always: Always use AI credits when the baseline quota is exhausted (will switch back automatically to using the baseline quota once the refresh hits) Baseline quota usage across models can be viewed in the settings page. Other ----- There is currently no support for: * Bring-your-own-key or bring-your-own-endpoint for additional rate limits * Organizational tiers in general availability, or via contract On this Page --- # Google Antigravity Documentation * side\_navigation * [Editor](https://antigravity.google/docs/editor) \>* Review Changes + Source Control Review Changes + Source Control =============================== Once the agent has begun writing code within a conversation, you’ll see a `Review Changes` section within the Agent panel’s [bottom toolbar](https://antigravity.google/docs/docs/agent-side-panel) . Clicking it will open up a pane within your editor where you can scroll through all of the changes you and your agent made within the conversation. ![Editor Review Changes](https://antigravity.google/docs/assets/image/docs/editor/review_changes_editor.png) Just like with artifacts, you can comment on any of the file diffs to communicate with the agent. ![Editor Source Control](https://antigravity.google/docs/assets/image/docs/editor/source_control_editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Editor](https://antigravity.google/docs/editor) \>* Agent Side Panel Agent Side Panel ================ Use the panel on the right side of the editor to work directly with the agent. You can spin up new conversations, attach images, switch [agent modes](https://antigravity.google/docs/docs/agent-modes-settings) , and select between [different models](https://antigravity.google/docs/docs/models) . ![Editor Agent Panel](https://antigravity.google/docs/assets/image/docs/editor/editor_agent_panel.png) As your conversation progresses, you’ll be able to keep track of open file changes, running terminal processes, and artifacts in the bottom toolbar above the input. ![Editor Agent Panel](https://antigravity.google/docs/assets/image/docs/editor/agent_panel_toolbar.png) On this Page --- # Google Antigravity Documentation * side\_navigation * Editor Editor ====== The primary entry point to Antigravity is our Editor, a surface based upon the VS Code codebase but full of rich AI-enabled features designed to improve your code-writing experience. Much of this editor is designed to feel the same as prior experiences you may have had. You can open files up, tab through them, edit them directly, get suggestions with [Tab](https://antigravity.google/docs/docs/tab) , and work with an agent on [smaller](https://antigravity.google/docs/docs/command) or [larger tasks](https://antigravity.google/docs/docs/agent-side-panel) . When you’re done, you can review your changes and interface with your preferred [source control](https://antigravity.google/docs/docs/review-changes-editor) . You can also still download extensions from the Open VSX marketplace to augment your experience, through further syntax highlighting, source control integrations, or other additions. ![Editor Screenshot](https://antigravity.google/docs/assets/image/docs/editor/editor.png) On this Page --- # Google Antigravity Documentation * side\_navigation * [Agent Manager](https://antigravity.google/docs/agent-manager) \>* [Conversation View](https://antigravity.google/docs/conversation-view) \>* Terminal Terminal ======== The agent manager window has terminal support as well! To toggle this, use Cmd/Ctrl + J to open the bottom pane of the agent manager, which is where terminals live. They are attached to the workspace that your current conversation is in. **Note:** Agent manager window's terminal integration works for local workspaces only, and Agent-used terminals run inside the editor window. ![agent-manager-terminal](https://antigravity.google/docs/assets/image/docs/agent-manager-terminal.png) On this Page ---