antigravity-mcp-bridge

antigravity-mcp-bridge

MCP server for Antigravity IDE that registers local tools and provides an OpenAI-compatible API bridge with real-time database context.

Category
访问服务器

README

Antigravity MCP Server & OpenAI-Compatible API Bridge

This project provides a Dual-Mode Integration solution for Google Antigravity:

  1. OpenAI-Compatible REST API Bridge (server.js): A local HTTP REST API server (http://localhost:8088/v1) that enables WhatsApp bots, Node.js scripts, Python clients, or any external application using the openai SDK (const { OpenAI } = require('openai')) to communicate with Antigravity models and automatically integrate with real-time Widyartha database context.
  2. MCP Stdio Server (index.js): A local Model Context Protocol (Stdio) server for registering local tools directly inside the Antigravity IDE UI without requiring external API keys.

🛠️ Key Features

  • OpenAI SDK Integration: Seamlessly use standard openai libraries by redirecting baseURL to http://localhost:8088/v1.
  • Real-Time MCP Data Integration: The HTTP server automatically enriches AI completion prompts with product catalog stock, order status, and invoice information from the local Widyartha SQLite database.
  • Comprehensive Model Selection: Supports model presets matching the Antigravity IDE UI dropdown:
    • gemini-3.6-flash-high, gemini-3.6-flash-medium, gemini-3.6-flash-low
    • gemini-3.5-flash-high, gemini-3.5-flash-medium, gemini-3.5-flash-low
    • gemini-3.1-pro-high, gemini-3.1-pro-low
    • claude-sonnet-4.6, claude-opus-4.6
    • gpt-oss-120b, antigravity
  • Auto-Setup & Portable: Automatically generates local security tokens, auto-installs missing dependencies, and uses relative paths for full portability across drives/folders.

🚀 Running the API Server

1. Start the HTTP Server (Port 8088)

Run the following command in your terminal:

npm start

Or double-click start-server.bat (Windows) / start-server.sh (Linux/macOS).

The server will listen at:

  • Base URL: http://localhost:8088/v1
  • Chat Completion: POST http://localhost:8088/v1/chat/completions
  • List Models: GET http://localhost:8088/v1/models

💻 Client Usage Examples (Node.js & Python)

Node.js Example:

const { OpenAI } = require('openai');

const openai = new OpenAI({
  apiKey: 'sk-random-key',
  baseURL: 'http://localhost:8088/v1'
});

async function main() {
  const response = await openai.chat.completions.create({
    model: 'antigravity',
    messages: [
      { role: 'user', content: 'Check batik shirt stock' }
    ]
  });

  console.log(response.choices[0].message.content);
}

main();

Python Example:

from openai import OpenAI

client = OpenAI(
    api_key="sk-random-key",
    base_url="http://localhost:8088/v1"
)

response = client.chat.completions.create(
    model="antigravity",
    messages=[
        {"role": "user", "content": "Check batik shirt stock"}
    ]
)

print(response.choices[0].message.content)

📋 Installing MCP Server in Antigravity IDE UI

To register the MCP Server directly in the Antigravity IDE UI:

  1. Open the Customizations panel in Antigravity IDE.
  2. Click the Open MCP Config button.
  3. Paste the following JSON configuration:
{
  "mcpServers": {
    "antigravity-mcp": {
      "command": "node",
      "args": [
        "E:/code/ai-bot-cs/MCP/antigravity/index.js"
      ],
      "cwd": "E:/code/ai-bot-cs/MCP/antigravity"
    }
  }
}
  1. Save the file and click Refresh 🔄.

🧪 Testing (Test Scripts)

All test scripts are located inside the test/ directory:

  • Run Node.js OpenAI SDK Test:

    npm test
    

    (Executes node test/client-example.js)

  • Run Python OpenAI Client Test:

    npm run test:python
    # Or: python test/test_openai_client.py
    
  • Run MCP Stdio Protocol Test:

    npm run test:mcp
    # Or: python test/check_mcp.py
    

📁 Directory Structure

E:\code\ai-bot-cs\MCP\antigravity\
├── server.js              # OpenAI Bridge HTTP REST API Server (Port 8088)
├── index.js               # MCP Stdio Server for Antigravity IDE UI
├── mcp-config.json        # Antigravity MCP Configuration Template
├── start-server.bat       # Portable Windows Launcher (Auto npm install)
├── start-server.sh        # Portable Linux/macOS Launcher
├── .env & .env.example    # Environment Variables
├── .gitignore             # Git Ignore Rules
├── package.json           # Package Manifest & Scripts
├── README.md              # English Documentation
└── test/                  # Test Scripts Directory
    ├── client-example.js     # Node.js OpenAI Client Test
    ├── test_openai_client.py # Python OpenAI Client Test
    └── check_mcp.py          # MCP Stdio Protocol Test

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选