Utopia

What altinoren/utopia does

altinoren/utopia is a simulation environment for experimenting with AI agents and household systems. Instead of connecting to physical appliances, it presents modeled devices and services through MCP tools. The available scenarios cover common home activities such as changing room conditions, operating lights and blinds, managing audio, checking food inventory, and securing the front door.

The scope also extends beyond home automation. The server includes a simulated autonomous electric vehicle, basic banking and credit-card operations, and grocery shopping based on previous purchases. These endpoints let developers test whether an agent can identify relevant tools and combine their results to complete a user request.

How it works

The agent discovers the exposed MCP tools and resources, then selects and calls them as needed. The Rooms resource supplies the house’s room list for operations that require a room, while Locations provides known places for the vehicle tool. This structure supports workflows that span several devices instead of requiring a separately programmed workflow for every request.

For example, a bedtime instruction can lead an agent to close blinds, adjust the bedroom temperature, lock the front door, turn off lights and audio, and check the vacuum state. The README presents this as an example of multi-endpoint orchestration performed by the agent. Results are simulated rather than applied to real equipment.

Setup and configuration

altinoren/utopia requires .NET 9 to build and run.

{
  "mcpServers": {
    "Utopia": {
      "command": "dotnet",
      "args": [
        "run",
        "--project",
        "<path_to_project_folder>/Utopia.csproj"
      ]
    }
  }
}

Replace the project path with the location of Utopia.csproj on the host running the server.

Tools and capabilities

The server currently lists these MCP tools:

• HumidityControl and AirQualityControl for room conditions.
• Lighting, MultiRoomAudio, SmartBlinds, and SmartThermostat for room devices.
• Refrigerator for viewing simulated contents.
• RobotVacuum for cleaning state and scheduling.
• SmartBed for sleep monitoring and adjustable climate control.
• SmartLock for the front door.
• AutonomousElectricVehicle for driving, charging, trip scheduling, and auto-parking.
• ACMEBank for basic accounts and credit-card operations.
• ACMEGrocery for shopping scenarios using past shopping data.

The refrigerator output can be combined with grocery operations, and room and location resources provide inputs for other tools. altinoren/utopia is therefore suited to testing tool selection, resource use, and chained actions.

Limitations and notes

The listed future ideas—pickup drones, garden robots, robotic kitchens, and medication assistance—are not described as currently available tools. Agent results can also depend on the client and model’s ability to reason across multiple available endpoints.