Antigravity integration
Google Antigravity is Google's agentic development platform, available as a desktop app and as the agy command-line tool. The same Apify plugin for Antigravity works in both and connects them to Actors in Apify Store. The plugin bundles:
- The Apify MCP server for searching Apify Store, running Actors, and retrieving datasets through the Model Context Protocol (MCP).
- An
apifyrouting agent that picks the right tool or skill from a natural-language request. - An always-on
apifyrule with the same routing guidance, so every session knows when to use Apify. - Five built-in skills for common workflows.
This integration uses a third-party service. If you find outdated content, please submit an issue on GitHub.
Prerequisites
- An Apify account - sign up for free if you don't have one.
- Antigravity - the desktop app, or the Antigravity CLI (
agy) version 1.4.0 or newer, installed and signed in. - Gemini Enterprise AI developer tools settings - only if your organization manages MCP servers, where admins must allow remote servers such as
https://mcp.apify.com.
Install the plugin
Install the plugin with the agy CLI, or by adding it to the Antigravity app plugins folder.
Antigravity CLI
-
Install the plugin from its repository:
agy plugin install https://github.com/apify/apify-antigravity-plugin -
Confirm the install:
agy plugin listThe
apifyplugin is listed as enabled, with its skills, agent, rule, and MCP server.
You can also install from the release ZIP: extract it into ~/.gemini/config/plugins, the folder the CLI and the app share.
Antigravity app
The app loads plugins from a folder on your machine.
-
Download
apify-antigravity-plugin.zipfrom the latest release and extract it into your global plugins folder, so the plugin ends up in~/.gemini/config/plugins/apify.If you have Git, you can clone the repository instead, which makes updates a
git pull:git clone https://github.com/apify/apify-antigravity-plugin ~/.gemini/config/plugins/apifyTo enable the plugin for a single project only, put the
apifyfolder in.agents/plugins/inside that project instead. These are the default plugin folders at the time of writing; check the current locations in the Antigravity plugins guide. -
Open Customizations > Installed and make sure Apify is listed under Plugins. It shows Setup required until you sign in.
Authenticate to Apify
The plugin bundles the Apify MCP server, which signs you in with OAuth. No API token goes into any configuration file.
- CLI: run
/mcp, select the Apify server, and select Authenticate. - App: open Customizations > Installed > Apify, select the Accounts tab, and select Sign in next to the
apifyaccount.
Antigravity opens a browser tab for the Apify OAuth flow, which ends with an authorization code. Copy the code, then paste it at the CLI prompt and press Enter, or paste it into the app and select Submit. The connection stays authenticated for future sessions, and you can revoke access at any time in Apify Console > Settings > Integrations.
Run your first prompt
Describe what you want in natural language. The apify rule and agent route the request to the right tool or skill, so you don't need to name tools yourself. To talk to the agent directly, start the CLI with agy --agent apify, or select apify from the app's Agents menu.
Use Apify to find a good Actor for scraping Google Maps places. Show me the best option, its input requirements, pricing model, and what kind of dataset output it returns. Do not run the Actor yet.
The agent searches Apify Store, fetches the top Actor's details through the Apify MCP server, and summarizes its inputs, pricing, and output - all without running the Actor.
Bundled skills
| Skill | Description |
|---|---|
apify-ultimate-scraper | Extraction using existing Actors for multi-step scraping and lead-generation workflows. |
apify-actor-development | Full Actor lifecycle - template selection, development, local testing, and deployment with apify push. |
apify-actorization | Converts existing JavaScript, TypeScript, Python, or CLI projects into Apify Actors. |
apify-generate-output-schema | Generates dataset and key-value store schemas for existing Actors. |
apify-sdk-integration | Integrates Actor execution into applications using the apify-client package. |
The agent invokes these skills automatically. Each skill is also a slash command, for example /apify-ultimate-scraper.
To run apify-ultimate-scraper:
Find 10 highly rated coffee shops in Seattle with name, address, rating, phone, and website.To run apify-actor-development:
Create an Apify Actor that accepts a `startUrl` and `maxPages` input, crawls the site, and stores each page title and URL.To run apify-sdk-integration:
Add Apify to this project. The Node.js API route should run an Actor and return dataset items as JSON.The Actor development, actorization, and scraping skills call the Apify CLI. Install it with npm install -g apify-cli and run apify login before you use them.
Troubleshooting
Most problems are a plugin that did not load, the wrong skill, the terminal sandbox, or OAuth on a machine without a browser.
The Apify MCP server doesn't appear
Restart the CLI or the app after installing the plugin, then run /mcp in the CLI or open Customizations > Installed > Apify > MCP in the app. If the server is listed but not connected, authenticate it again; see Authenticate to Apify.
Antigravity picks the wrong skill
Start the CLI with agy --agent apify, or select apify from the app's Agents menu, so the routing agent handles the request. The agent owns the guardrails that pick the right skill and avoid common traps, such as confusing the apify and apify-client packages.
Commands that need the network hang or fail
With the default permission preset, Antigravity runs agent commands in a terminal sandbox without network access. Commands that need the network, such as npm install or Apify CLI commands, retry and then fail. The Apify CLI then reports that it couldn't verify your API token, even though your login is fine. Let the agent run the command outside the sandbox and approve the request, or add the domains the command needs to your read_url permissions, which Antigravity adds to the sandbox's outbound allowlist.
Browser doesn't open, or OAuth fails
The MCP sign-in doesn't need a browser on the machine that runs Antigravity. Over SSH or in a container, open the sign-in link that Antigravity shows in a browser on any machine, then paste the authorization code back at the agy prompt.
If you can't use OAuth at all, use the skills that don't need the MCP server and authenticate them with your API token. Copy it from Apify Console > Settings > Integrations and set it before starting Antigravity:
export APIFY_TOKEN=<YOUR_API_TOKEN>
The SDK integration skill uses the apify-client package, which reads APIFY_TOKEN from your environment. The Actor development, actorization, and scraping skills use the Apify CLI, which reads APIFY_TOKEN from version 1.11.0. On older versions, log in with the token once:
apify login --token "$APIFY_TOKEN"
Limitations
- Long-running Actors may exceed the time a single tool call waits for completion. Reduce the scope or split the work across multiple prompts.
- Each Actor run consumes Apify platform usage from your plan in addition to any Antigravity usage. See Billing for details.
- Skills that edit files in your project (Actor development, actorization, and SDK integration) make local changes. Review them before deploying or committing.
Related integrations
- Claude Code CLI integration - Install the Apify plugin for Claude Code CLI
- MCP server integration - Use the Apify MCP server with other clients
Resources
- Apify plugin for Antigravity - Source repository and README with advanced setup notes
- Antigravity plugins guide - Official docs for managing Antigravity plugins
- Apify Store - Browse Actors you can run from Antigravity