About Computer Use
Computer Use is an MCP server built with FastMCP and Orgo for launching and controlling cloud virtual desktops. It serves computer-use agents with mouse, keyboard, screenshot, and terminal controls, and can invoke Claude computer use.
Key features
- 01Start and manage virtual computer sessions
- 02Capture screenshots
- 03Control mouse and keyboard
- 04Run terminal commands
- 05Invoke the Claude computer-use prompt
Use cases
- 01Let an agent operate apps in a virtual desktop
- 02Capture desktop images for task inspection
- 03Run terminal commands in a cloud environment
MCP support
The MCP server exposes tools for initializing virtual computers, screenshots, input control, command execution, and session management. It also includes a prompt tool for Claude computer use.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEComputer Use MCP Server 🖥️
Control virtual computers through the Model Context Protocol (MCP). Built with FastMCP and Orgo.
Spin up cloud desktop environments for Computer Use Agents (CUA).
Quick Start
1. Install Dependencies
# Create virtual environment
python3 -m venv mcp-env
source mcp-env/bin/activate
# Install packages
pip install fastmcp pydantic orgo
2. Get API Keys
- Orgo API Key: Sign up at orgo.ai
- Anthropic API Key: Get from console.anthropic.com
3. Run Server
# Set environment variables
export ORGO_API_KEY="your_orgo_key"
export ANTHROPIC_API_KEY="your_anthropic_key"
# Start server
python computer_mcp_server.py
Server runs on http://127.0.0.1:9000
Test Client
import asyncio
from fastmcp import Client
async def demo():
async with Client("http://127.0.0.1:9000/mcp") as client:
# Initialize computer
result = await client.call_tool("initialize_computer", {
"api_key": "your_orgo_key"
})
print(f"Computer ready: {result[0].text}")
# Take screenshot
screenshot = await client.call_tool("get_screenshot")
print("Screenshot taken!")
# Click and type
await client.call_tool("left_click", {"x": 100, "y": 200})
await client.call_tool("type_text", {"text": "Hello World"})
await client.call_tool("press_key", {"key": "Enter"})
if __name__ == "__main__":
asyncio.run(demo())
Available Tools
| Tool | Description |
|---|---|
initialize_computer | Start virtual computer |
get_screenshot | Take screen capture |
left_click, right_click, double_click | Mouse actions |
scroll | Scroll up/down |
type_text, press_key | Keyboard input |
wait | Wait for specified seconds |
execute_bash | Run terminal commands |
restart_computer, shutdown_computer | Computer lifecycle |
get_status, list_sessions | Status and session management |
prompt | Claude AI Computer Use |
Claude AI Computer Use Example
# Let Claude control the computer
await client.call_tool("prompt", {
"instruction": "Open Firefox and go to google.com"
})
Requirements
- Python 3.8+
- Orgo API key
- Anthropic API key (for Claude CUA)
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.