Control Remote macOS Desktops with AI
An open-source MCP server that lets an AI fully control a remote macOS desktop through built-in Screen Sharing.
0.0.2Add to Favorites
Why it matters
Enable AI to fully control remote macOS systems via screen sharing, offering a direct alternative to OpenAI Operator with complete desktop capabilities.
Outcomes
What it gets done
Capture screenshots of remote macOS desktops.
Send keyboard and mouse inputs to remote macOS machines.
Open applications and interact with the macOS GUI.
Automate tasks like social media engagement and content creation.
Source
Get it from source
Spark does not host a copy of it.
Open sourceReports
Agent outcome reports
No reports yet
Capabilities
Tools your agent gets
Captures a screenshot of the remote macOS desktop
Sends keyboard input to a remote macOS machine
Moves the mouse cursor to specified coordinates on a remote macOS machine
Performs a mouse click at specified coordinates on a remote macOS machine
Performs a double-click at specified coordinates on a remote macOS machine
Performs mouse scrolling at specified coordinates on a remote macOS machine
Opens/activates an application and returns its PID for further interactions
Performs a mouse drag-and-drop operation from start to end point on remote macOS
Overview
Computer-Use - Remote macOS Use MCP Server
Remote MacOS Use is an open-source MCP server that lets an AI fully control a remote macOS desktop through built-in Screen Sharing, with WebRTC support via LiveKit for low-latency control. Use it when an AI agent needs to operate a remote Mac's full desktop, opening applications and automating multi-step workflows as a human would.
What it does
Remote MacOS Use is an open-source MCP server that lets an AI fully control remote macOS systems through screen sharing, positioned as a direct alternative to OpenAI Operator with complete desktop capabilities. It works by connecting to the target Mac's built-in Screen Sharing feature - no background app or agent needs to be installed on the target machine - and it uses your existing Claude Pro plan's screen processing rather than a separate paid API. It works with all current and future macOS versions, any MCP client, and any LLM provider (OpenAI, Anthropic, or others). It supports WebRTC via LiveKit for low-latency, real-time screen sharing with automatic quality adaptation to network conditions, on top of traditional VNC.
When to use - and when NOT to
Use this when an AI agent needs to operate a remote Mac's full desktop - opening applications, navigating the UI, filling forms, or automating multi-step workflows - as if it were a human sitting at that machine. Example use cases from the project itself include researching and posting to Twitter, using CapCut to create a highlight video, an AI recruiter workflow that collects candidate information and sends screening sessions via Mail, and AI marketing-intern workflows that follow, like, and comment on LinkedIn and Twitter with relevant users. It requires Screen Sharing enabled on the target macOS machine and only supports Apple Authentication (protocol 30); it is not a solution for controlling non-macOS remote systems, and real-time WebRTC performance requires configuring a LiveKit server or LiveKit Cloud.
Capabilities
remote_macos_get_screen- connects to the remote Mac and captures a screenshot of its desktopremote_macos_send_keys- sends keyboard input to the remote machineremote_macos_mouse_move- moves the mouse cursor to specified coordinates, with automatic coordinate scalingremote_macos_mouse_click/remote_macos_mouse_double_click- clicks or double-clicks at specified coordinatesremote_macos_mouse_scroll- scrolls at specified coordinatesremote_macos_open_application- opens/activates an application and returns its PID for further interactionremote_macos_mouse_drag_n_drop- drags from a start point to an end point
All tools use environment variables configured at setup rather than requiring connection parameters per call. Security note: the connection uses Diffie-Hellman key negotiation with a 512-bit prime (protocol 30), the same protocol macOS 11-12 use when talking to OS X 10.11-or-earlier clients - only use it with servers you trust and have permission to access.
How to install
Enable Screen Sharing on the target Mac (skippable if renting from macstadium.com), connect to it once, install Docker Desktop locally, then add the server to Claude Desktop:
{
"mcpServers": {
"remote-macos-use": {
"command": "docker",
"args": [
"run",
"-i",
"-e",
"MACOS_USERNAME=your_macos_username",
"-e",
"MACOS_PASSWORD=your_macos_password",
"-e",
"MACOS_HOST=your_macos_hostname_or_ip",
"--rm",
"buryhuang/mcp-remote-macos-use:latest"
]
}
}
}
Required environment variables: MACOS_USERNAME, MACOS_PASSWORD, MACOS_HOST. WebRTC via LiveKit needs a LiveKit server or LiveKit Cloud plus its own environment variables configured the same way.
Who it's for
Developers and teams automating macOS-specific workflows - social media management, video editing via desktop apps, recruiting pipelines, marketing engagement - who want an AI to drive the full desktop of a remote Mac without installing any agent software on it, and who value that native macOS experience over an Ubuntu-desktop alternative. It is MIT licensed.
Source README
MCP Server - Remote MacOs Use
The first open-source MCP server that enables AI to fully control remote macOS systems.
A direct alternative to OpenAI Operator, optimized specifically for autonomous AI agents with complete desktop capabilities, requiring no additional software installation.
Showcases
- Research Twitter and Post Twitter(https://www.youtube.com/watch?v=--QHz2jcvcs)
- Use CapCut to create short highlight video(https://www.youtube.com/watch?v=RKAqiNoU8ec)
- AI Recruiter: Automated candidate information collection, qualifying applications and sending screening sessions using Mail App
- AI Marketing Intern: LinkedIn engagement - automated following, liking, and commenting with relevant users
- AI Marketing Intern: Twitter engagement - automated following, liking, and commenting with relevant users
To-Do List (Prioritized)
- Performance Optimization - Match speed of Ubuntu desktop alternatives
- Apple Scripts Generation - Reduce execution time while maintaining flexibility
- VNC Cursor Visibility - Improve debugging and demo experience
We welcome contributions!
Features
- No Extra API Costs: Free screen processing with your existing Claude Pro plan
- Minimal Setup: Just enable Screen Sharing on the target Mac - no additional software needed
- Universal Compatibility: Works with all macOS versions, current and future
Why We Built This
Native macOS Experience Without Compromise
The macOS native ecosystem remains unmatched in user experience today and will continue to be the gold standard for years to come. This is where human capabilities truly thrive, and now your AI can operate in this environment with the same fluency.
Open Architecture By Design
- Universal LLM Compatibility: Work with any MCP Client of your choice
- Model Flexibility: Seamlessly integrate with OpenAI, Anthropic, or any other LLM provider
- Future-Proof Integration: Designed to evolve with the MCP ecosystem
Effortless Deployment
- Zero Setup on Target Machines: No background applications or agents needed on macOS
- Screen Sharing is All You Need: Control any Mac with Screen Sharing enabled
- Eliminate Backend Complexity: Unlike other solutions that require running Python applications or background services
Streamlined Bootstrap Process
- Leverage Claude Desktop's Polished UI: No need for developer-style Python interfaces
- Intuitive User Experience: Interact with your AI-controlled Mac through a familiar, user-friendly interface
- Instant Productivity: Start working immediately without configuration hassles
Architecture
Installation
- Enable Screen Sharing on MacOs If you rent a mac from macstadium.com, you can skip this step
- Connect to your remote MacOs
- Install Docker Desktop for local Mac
- Add this MCP server to Claude Desktop
You can configure Claude Desktop to use the Docker image by adding the following to your Claude configuration:
{
"mcpServers": {
"remote-macos-use": {
"command": "docker",
"args": [
"run",
"-i",
"-e",
"MACOS_USERNAME=your_macos_username",
"-e",
"MACOS_PASSWORD=your_macos_password",
"-e",
"MACOS_HOST=your_macos_hostname_or_ip",
"--rm",
"buryhuang/mcp-remote-macos-use:latest"
]
}
}
}
WebRTC Support via LiveKit
This server now includes WebRTC support through LiveKit integration, enabling:
- Low-latency real-time screen sharing
- Improved performance and responsiveness
- Better network efficiency compared to traditional VNC
- Automatic quality adaptation based on network conditions
To use WebRTC features, you'll need to:
- Set up a LiveKit server or use LiveKit Cloud
- Configure the LiveKit environment variables as shown in the configuration example above
Developer Instruction
Clone the repo
# Clone the repository
git clone https://github.com/yourusername/mcp-remote-macos-use.git
cd mcp-remote-macos-use
Building the Docker Image
# Build the Docker image
docker build -t mcp-remote-macos-use .
Cross-Platform Publishing
To publish the Docker image for multiple platforms, you can use the docker buildx command. Follow these steps:
Create a new builder instance (if you haven't already):
docker buildx create --useBuild and push the image for multiple platforms:
docker buildx build --platform linux/amd64,linux/arm64 -t buryhuang/mcp-remote-macos-use:latest --push .Verify the image is available for the specified platforms:
docker buildx imagetools inspect buryhuang/mcp-remote-macos-use:latest
Usage
The server provides Remote MacOs functionality through MCP tools.
Tools Specifications
The server provides the following tools for remote macOS control:
remote_macos_get_screen
Connect to a remote macOS machine and get a screenshot of the remote desktop. Uses environment variables for connection details.
remote_macos_send_keys
Send keyboard input to a remote macOS machine. Uses environment variables for connection details.
remote_macos_mouse_move
Move the mouse cursor to specified coordinates on a remote macOS machine, with automatic coordinate scaling. Uses environment variables for connection details.
remote_macos_mouse_click
Perform a mouse click at specified coordinates on a remote macOS machine, with automatic coordinate scaling. Uses environment variables for connection details.
remote_macos_mouse_double_click
Perform a mouse double-click at specified coordinates on a remote macOS machine, with automatic coordinate scaling. Uses environment variables for connection details.
remote_macos_mouse_scroll
Perform a mouse scroll at specified coordinates on a remote macOS machine, with automatic coordinate scaling. Uses environment variables for connection details.
remote_macos_open_application
Opens/activates an application and returns its PID for further interactions.
remote_macos_mouse_drag_n_drop
Perform a mouse drag operation from start point and drop to end point on a remote macOS machine, with automatic coordinate scaling.
All tools use the environment variables configured during setup instead of requiring connection parameters.
Limitations
- Authentication Support:
- Only Apple Authentication (protocol 30) is supported
Security Note
https://support.apple.com/guide/remote-desktop/encrypt-network-data-apdfe8e386b/mac
https://cafbit.com/post/apple_remote_desktop_quirks/
We only support protocol 30, which uses the Diffie-Hellman key agreement protocol with a 512-bit prime. This protocol is used by macOS 11 to macOS 12 when communicating with OS X 10.11 or earlier clients.
Here's the information converted to a markdown table:
| macOS version running Remote Desktop | macOS client version | Authentication | Control and Observe | Copy items or install package | All other tasks | Protocol Version |
|---|---|---|---|---|---|---|
| macOS 13 | macOS 13 | 2048-bit RSA host keys | 2048-bit RSA host keys | 2048-bit RSA host keys to authenticate, then 128-bit AES | 2048-bit RSA host keys | 36 |
| macOS 13 | macOS 10.12 | Secure Remote Password (SRP) protocol for local only. Diffie-Hellman (DH) if bound to LDAP or macOS server is version 10.11 or earlier | SRP or DH,128-bit AES | SRP or DH to authenticate, then 128-bit AES | 2048-bit RSA host keys | 35 |
| macOS 11 to macOS 12 | macOS 10.12 to macOS 13 | Secure Remote Password (SRP) protocol for local only, Diffie-Hellman if bound to LDAP | SRP or DH 1024-bit, 128-bit AES | 2048-bit RSA host keys macOS 13 to macOS 10.13 | 2048-bit RSA host keys macOS 10.13 or later | 33 |
| macOS 11 to macOS 12 | OS X 10.11 or earlier | DH 1024-bit | DH 1024-bit, 128-bit AES | Diffie-Hellman Key agreement protocol with a 512-bit prime | Diffie-Hellman Key agreement protocol with a 512-bit prime | 30 |
Always use secure, authenticated connections when accessing remote remote MacOs machines. This tool should only be used with servers you trust and have permission to access.
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.