Antigravity MCP Bridge
Antigravity 子代理桥接(FastMCP)
一个容器化的 FastMCP 服务器,将 Google Antigravity CLI(agy --headless) 作为远程模型上下文协议(MCP)工具通过 SSE 暴露出来。
该桥接允许对话式 AI 客户端(如 Google Spark、Claude、Cursor 或 ChatGPT)在沙盒化、容器化的工作区中调度完整的编码和自动化子代理。
🚀 主要特性
**同步子代理执行(
execute_antigravity_sync)**: 将编码/自动化任务直接分派给agy --headless`,并在代理完成后返回输出日志(适用于 1-3 分钟的短任务)。异步后台执行(
start_antigravity_async): 在后台生成一个长时间运行的 Antigravity 代理,并返回一个task_id以实现解耦执行。任务轮询与管理(
get_antigravity_task_status、list_antigravity_tasks): 轮询进度、检查日志并列出所有后台子代理。工作区隔离与沙盒化: 在 Docker 容器中运行,仅挂载
./workspaces,保护主机系统文件,同时允许代理在其沙盒内完全自主地进行构建、测试和执行。与
mcp-server-oauth-proxy无缝集成: 在8080/sse端口暴露标准 SSE 端点,可直接插入 Home Assistant 的 OAuth 代理,以便从 Google Spark 安全访问。
Related MCP server: agy-mcp
🛠️ 仓库结构
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📦 快速开始与部署
1. 配置环境变量
复制 .env.example 模板:
cp .env.example .env(可选:如果您的 agy 环境需要,请添加 GEMINI_API_KEY 或 ANTIGRAVITY_API_KEY。)
2. 使用 Docker Compose 启动
docker compose up -d --buildFastMCP 服务器将开始监听:
👉 http://localhost:8080/sse(或您的主机局域网 IP)。
3. 验证容器健康状态
docker compose logs -f🔗 连接到 mcp-server-oauth-proxy
要通过您的 Home Assistant OAuth 代理将此 Antigravity 子代理桥接安全地暴露给 Google Spark:
在 Home Assistant 中,打开 设置 > 加载项 > MCP Server OAuth Proxy > 配置。
在
servers下添加一个新的服务器条目:
- 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: ""保存并重启加载项。
在 Google Spark 中,添加自定义工具,指向
https://antigravity-oauth.yourdomain.com/sse,客户端 ID 为spark-antigravity。
🧰 可用的 MCP 工具
1. execute_antigravity_sync
描述:通过
agy --headless同步执行任务并等待输出。参数:
prompt(字符串,必填):目标或编码指令。sub_path(字符串,可选):在/workspace中执行的子目录。timeout_seconds(整数,默认值:300):中止前等待的最大秒数。
2. start_antigravity_async
描述:在后台启动一个 Antigravity 任务并返回
task_id。参数:
prompt(字符串,必填):目标或编码指令。sub_path(字符串,可选):在/workspace中执行的子目录。
3. get_antigravity_task_status
描述:检查后台任务的进度和完整日志。
参数:
task_id(字符串,必填):由start_antigravity_async返回的 8 字符 ID。
4. list_antigravity_tasks
描述:列出所有活动及已完成的后台任务。
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceRuns isolated AI agents in Docker containers with persistent workspaces and conversation history. Each agent has access to file operations, shell commands, GitHub integration, and can connect to external MCP servers for additional tools.3MIT
- AlicenseAqualityCmaintenanceWraps Google Antigravity CLI into 11 typed MCP tools, enabling any MCP client to invoke agy for code review, prototyping, execution, and long-running tasks.1116MIT
- AlicenseBqualityAmaintenanceAn MCP bridge that lets Codex delegate long-running agent work to the Antigravity CLI, providing observable and resumable tool-based execution with project scoping.135MIT
- AlicenseNot gradedqualityAmaintenanceIntegrates Google Antigravity's agy agent into MCP, enabling asynchronous delegation of complex tasks like code analysis and file management with job polling.1002MIT
Related MCP Connectors
A paid remote MCP for OpenAI Codex agent coordination MCP, built to return verdicts, receipts, usage
OCR, transcription, file extraction, and image generation for AI agents via MCP.
Control plane for autonomous software labor. Agents claim objectives over MCP with audit trail.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/connormxy/antigravity-mcp-bridge'
If you have feedback or need assistance with the MCP directory API, please join our Discord server