nandhakumar-murugan/gemini-antigravity-bridge

The World's First Bidirectional MCP Bridge Between Google Gemini Spark & DeepMind Antigravity
**Connect Google's cloud AI (Gemini Spark powered by Gemini 3.8 / 3.7) to your local agentic IDE (Antigravity) with 24 powerful MCP tools, hardened by OpenAgentShield Zero-Trust runtime sandboxing and automatic credential redaction.**
Quick Start · Architecture · Tools · Connected Apps · Deploy
---
🌐 What Is This?
**Gemini Antigravity Bridge** is an MCP (Model Context Protocol) server that creates a **persistent, bidirectional communication channel** between:
- ☁️ **Google Gemini Spark** (cloud-based AI assistant with Google Workspace access, powered by Gemini 3.8 / 3.7)
- 💻 **Google DeepMind Antigravity** (local agentic IDE with full system access)
Optimized for **Gemini 3.8's next-generation agentic reasoning**, this enables a **fully autonomous loop** where Spark can dispatch coding tasks to your local machine, and Antigravity can request cloud intelligence back from Spark — all through a standardized, secure protocol.
---
🏆 Why This Bridge?
| Feature | This Project | Other MCP Servers |
|---|:---:|:---:|
| Bidirectional Cloud ↔ Local | ✅ | ❌ |
| Autonomous Agent Task Dispatch | ✅ | ❌ |
| Cross-Agent Shared Memory | ✅ | ❌ |
| Antigravity Conversation Injection | ✅ | ❌ |
| Spark Connected Apps Orchestration | ✅ | ❌ |
| Google Workspace Integration | ✅ (via Spark) | Partial |
| Local File System Access | ✅ | Some |
| System Command Execution | ✅ | Some |
---
🚀 Quick Start
Prerequisites
- Python 3.10+
- ngrok account (free tier) with a reserved domain
- Google Gemini Spark (Advanced/Ultra subscription)
- Google DeepMind Antigravity IDE
Installation
Connect to Gemini Spark
1. Open Gemini Spark → Settings → Connected Apps
2. Click **"Add a custom app"**
3. Enter your public MCP URL: https://your-domain.ngrok-free.dev/mcp
4. Spark will discover all 23 tools automatically!
---
🛠️ Available Tools (24)
🔧 System Execution
| Tool | Description |
|---|---|
| run_system_command | Execute shell/PowerShell command with AST risk analysis and captured output |
| run_batch_commands | Run multiple commands sequentially with error handling and policy validation |
📂 File Operations
| Tool | Description |
|---|---|
| read_file | Read file contents with sensitive credential protection |
| write_file | Create or overwrite files with automatic secret redaction |
| edit_file | Surgically edit specific lines in existing files |
| append_file | Append content to the end of a file |
| batch_write_files | Create multiple files in a single operation with policy enforcement |
| create_full_project | Scaffold an entire project directory structure |
| list_directory | List directory contents with metadata |
🤖 Agent Orchestration
| Tool | Description |
|---|---|
| run_agent_task | Launch an autonomous Antigravity coding subagent |
| get_agent_status | Check status and output of a running agent task |
| terminate_task | Kill a running agent task |
🛡️ Zero-Trust Security & Telemetry (OpenAgentShield)
| Tool | Description |
|---|---|
| get_security_audit_log | Retrieves OpenAgentShield Zero-Trust audit events, blocked attacks, and system risk scores (DOI: 10.5281/zenodo.22259022) |
🧠 Shared Memory & Sync
| Tool | Description |
|---|---|
| get_bridge_history | Retrieve full cross-client operation history with redacted credentials |