Antigravity MCP Bridge
Allows AI clients to dispatch coding and automation tasks to Google Antigravity (agy --headless), providing synchronous and asynchronous execution, task polling, and workspace isolation.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Antigravity MCP BridgeRun a coding task to build a Flask API skeleton."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Antigravity Sub-Agent Bridge (FastMCP)
A containerized FastMCP server that exposes the Google Antigravity CLI (agy --headless) as remote Model Context Protocol (MCP) tools over SSE.
This bridge allows conversational AI clients (such as Google Spark, Claude, Cursor, or ChatGPT) to dispatch full coding and automation sub-agents in a sandboxed, containerized workspace.
🚀 Key Features
Synchronous Sub-Agent Execution (
execute_antigravity_sync): Dispatches a coding/automation task directly toagy --headlessand returns the output log once the agent finishes (ideal for short 1–3 minute tasks).Asynchronous Background Execution (
start_antigravity_async): Spawns a long-running Antigravity agent in the background, returning atask_idfor decoupled execution.Task Polling & Management (
get_antigravity_task_status,list_antigravity_tasks): Polls progress, inspects logs, and lists all background sub-agents.Workspace Isolation & Sandboxing: Runs within a Docker container mounting only
./workspaces, protecting host system files while giving the agent full build, test, and execution autonomy inside its sandbox.Seamless Integration with
mcp-server-oauth-proxy: Exposes standard SSE endpoints on port8080/sse, pluggable directly into Home Assistant's OAuth proxy for secure access from Google Spark.
Related MCP server: opencode-mcp-bridge
🛠️ Repository Layout
antigravity-mcp-bridge/
├── Dockerfile # Container image with Python 3.12, Node.js, git, & agy CLI
├── docker-compose.yml # Service definitions with volume mounts & port bindings
├── requirements.txt # fastmcp, mcp[cli], pydantic
├── server.py # FastMCP Server exposing Antigravity agent tools
├── .env.example # Environment variable template
├── .gitignore # Ignored artifacts & workspace directories
└── README.md # Documentation📦 Quickstart & Deployment
1. Configure Environment
Copy the .env.example template:
cp .env.example .env(Optional: Add your GEMINI_API_KEY or ANTIGRAVITY_API_KEY if required by your agy environment).
2. Start with Docker Compose
docker compose up -d --buildThe FastMCP server will start listening on:
👉 http://localhost:8080/sse (or your host LAN IP).
3. Verify Container Health
docker compose logs -fđź”— Connecting to mcp-server-oauth-proxy
To expose this Antigravity sub-agent bridge securely to Google Spark via your Home Assistant OAuth proxy:
In Home Assistant, open Settings > Add-ons > MCP Server OAuth Proxy > Configuration.
Add a new server entry under
servers:
- client_name: "Google Spark"
server_name: "Antigravity Bridge"
server_url: "http://<YOUR_DOCKER_HOST_IP>:8080/sse"
bearer_token: ""
client_id: "spark-antigravity"
client_secret: "secret-antigravity"
allowed_redirect_uris: "https://oauth-redirect.googleusercontent.com"
host: "antigravity-oauth.yourdomain.com"
path_prefix: ""Save and restart the add-on.
In Google Spark, add the custom tool pointing to
https://antigravity-oauth.yourdomain.com/ssewith Client IDspark-antigravity.
đź§° Available MCP Tools
1. execute_antigravity_sync
Description: Executes a task synchronously via
agy --headlessand waits for output.Parameters:
prompt(string, required): The goal or coding instruction.sub_path(string, optional): Subdirectory in/workspaceto execute within.timeout_seconds(int, default: 300): Max seconds to wait before aborting.
2. start_antigravity_async
Description: Starts an Antigravity task in the background and returns a
task_id.Parameters:
prompt(string, required): The goal or coding instruction.sub_path(string, optional): Subdirectory in/workspaceto execute within.
3. get_antigravity_task_status
Description: Inspects progress and full logs of a background task.
Parameters:
task_id(string, required): The 8-character ID returned bystart_antigravity_async.
4. list_antigravity_tasks
Description: Lists all active and completed background tasks.
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
Discover and call AI agents via MCP. Supports A2A agents and platform agents with async tasks.
- DazbenchOAuthapp.dazbench
Task management your AI agents can actually run. One line becomes a context-ready task over MCP.
Your org's AI agents, tasks, runs, search, and brain files as MCP tools and resources.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceIntegrates Google Antigravity's agy agent into MCP, enabling asynchronous delegation of complex tasks like code analysis and file management with job polling.33 npm3MIT
- AlicenseNot gradedqualityBmaintenanceExposes opencode's coding agent and shell as MCP tools, enabling MCP-only AI clients to execute shell commands, manage files, and run agent sessions with async job handling.16 npm1MIT
- AlicenseAqualityCmaintenanceEnables AI harnesses to delegate code analysis, modification, testing, and long-running tasks to the locally installed Antigravity CLI via stdio MCP, with job status tracking and conversation continuity.3MIT
- AlicenseBqualityAmaintenanceEnables delegating coding tasks to the official Antigravity CLI from any MCP client, with task listing, model selection, streamed event tracking, cancellation, and resumable sessions. Runs each task against an isolated copy of the repository so results can be previewed, tested, and only integrated into the original after explicit user confirmation.2363 npmMIT