claude-antigravity-mcp-bridge
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., "@claude-antigravity-mcp-bridgeUse the Antigravity CLI to generate a haiku about the ocean."
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.
Claude Antigravity MCP Bridge
A Model Context Protocol (MCP) bridge enabling Claude Desktop to execute tasks seamlessly via the Google Antigravity CLI (agy).
This bridge exposes the Antigravity CLI to Claude Desktop as a native tool, allowing Claude to harness all capabilities of the Antigravity ecosystem, including its models, skills, plugins, and background workers, completely free from the constraints of a standard chat interface.
Architecture
sequenceDiagram
participant Claude as Claude Desktop
participant MCP as claude-antigravity-mcp-bridge
participant AGY as Antigravity CLI (agy)
Claude->>MCP: Call tool `run_agy_task` with flags
MCP->>AGY: Spawns subprocess (stdin=DEVNULL)
AGY-->>MCP: Executes task & returns stdout/stderr
MCP-->>Claude: Tool resultRelated MCP server: agy-headless-bridge MCP server
Setup & Installation
Ensure you have Python 3.9+ installed and the Antigravity CLI available in your system path.
Clone the repository:
git clone https://github.com/mahdi/claude-antigravity-mcp-bridge.git cd claude-antigravity-mcp-bridgeCreate a virtual environment:
python -m venv .venv # Windows: .\.venv\Scripts\activate # macOS/Linux: source .venv/bin/activateInstall the package:
pip install .
Configuration in Claude Desktop
Update your claude_desktop_config.json (accessible via Claude Desktop > Settings > Developer > Edit Config) to point to the bridge:
{
"mcpServers": {
"claude-antigravity-bridge": {
"command": "C:\\path\\to\\claude-antigravity-mcp-bridge\\.venv\\Scripts\\python.exe",
"args": ["-m", "claude_antigravity_bridge"]
}
}
}Note: Replace C:\\path\\to\\ with the actual absolute path to the cloned repository. Restart Claude Desktop after making this change.
Usage
Once configured, simply instruct Claude Desktop:
"Run the agy bridge and use the default model to print 'Hello'."
Claude will automatically know how to format flags like --print and --model because the tool description embeds a comprehensive Antigravity cheatsheet.
Security Considerations
The bridge passes prompts directly to the
agyCLI. Be cautious of prompt injection.Interactive terminal prompts from
agyare forcefully skipped (--dangerously-skip-permissions) to prevent the MCP process from hanging. Ensure your local agents are configured with appropriate safety boundaries.
This server cannot be deployed
Maintenance
Related MCP Connectors
Use your Mac, Windows or Linux computer from ChatGPT, Claude or Codex: files, commands, documents.
Work on your own computers from Claude or ChatGPT: run commands, edit and search files.
Connect AI assistants to your approved computer for files, commands, browser and desktop tasks.
- ZionDeskOAuthapp.ziondesk
Connect ChatGPT, Claude, Cursor or VS Code to your Windows PC: files, terminal, browser, apps.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables Claude to drive the Antigravity agent by running prompts through the agy CLI non-interactively, supporting session management and optional sandboxed execution.51ISC
- AlicenseBqualityAmaintenanceEnables calling Google's Antigravity CLI (agy) headlessly through MCP, providing tools like agy_ask and agy_research for agents like Claude Code.2105 PyPI19MIT
- AlicenseAqualityAmaintenanceLets Codex delegate tasks to the Google Antigravity CLI (agy) with background job execution, live progress tracking, a visual dashboard, and structured error diagnostics.8MIT
- AlicenseAqualityCmaintenanceLets MCP clients such as Claude Code, Claude Desktop, Cursor, or Codex hand tasks to the Antigravity CLI (agy) so the work runs on the user's Antigravity plan instead of the client's. Exposes tools for delegated coding tasks and reviews, web search with sources, image/PDF analysis, shell command generation or execution, brainstorming, and model/quota checks.7MIT