MCP Server

The Emuluxe MCP Server is an implementation of the Model Context Protocol (MCP), allowing AI agents—like Claude Desktop, Cursor, and Antigravity—to programmatically control the Emuluxe hardware foundry.

By connecting this server, your AI assistant can instantly "see" your local code on physical hardware, perform automated UI audits, and verify mobile responsiveness without you leaving your editor.

[!TIP] Unified Ecosystem: Both the IDE extension and the MCP server share the same API Token. You can use the extension for your own visual previews while your AI agent uses the MCP server for parallel, background simulations.

Capabilities

The server exposes a core toolset that AI agents can utilize:

`simulate`

Initiates a high-fidelity hardware simulation for any URL.

url (Required): The target URL (e.g., http://localhost:3000).
device: The hardware profile (e.g., iphone-15-pro-max, pixel-8-pro).
browser: The engine to emulate (safari or chrome).

Setup & Configuration

To use the MCP server, you must have a valid Emuluxe API Token. Generate one in your Integrations Settings.

🚀 Zero-Install (Recommended)

No manual installation or cloning is required. You can add the server to your IDE instantly using npx.

1. Cursor (Native Support)

Open Cursor Settings (Cmd + Shift + J).
Go to Features → MCP Servers.
Click Add New MCP Server.
Name: Emuluxe
Type: command
Command:
```
npx -y @emuluxe/mcp-server
```
Environment Variables:
- EMULUXE_TOKEN: your_api_token_here
- EMULUXE_IDE: cursor

2. VS Code (via Roo Code / Cline)

Add this entry to your mcp_config.json:

"emuluxe": {
  "command": "npx",
  "args": ["-y", "@emuluxe/mcp-server"],
  "env": {
    "EMULUXE_TOKEN": "your_api_token_here"
  }
}

3. Antigravity

Add this entry to your .gemini/antigravity/mcp_config.json:

"emuluxe": {
  "command": "npx",
  "args": ["-y", "@emuluxe/mcp-server"],
  "env": {
    "EMULUXE_TOKEN": "your_api_token_here"
  }
}

Usage Example

Once configured, you can simply ask your AI agent to view your work:

User Prompt:

"I've just updated the mobile navigation. Can you check how it looks on an iPhone 15 Pro Max at http://localhost:3000?"

AI Response:

"I've started a simulation for you. You can view it directly in your IDE Extension (recommended) or on the Web App if the extension is not installed: Launch Simulation"

Why use MCP?

AI-Aware Debugging: Your agent can trigger simulations to verify its own code changes.
Headless Audits: Connect your agent to the REST API and MCP to automate complex cross-device workflows.
Zero Context Switch: Stay in your IDE while the agent handles the hardware orchestration.

Next, explore the CLI Toolchain.

Loading Emuluxe