Unreal MCP Server
Unreal Engine project automation through natural language actor and blueprint control.
Data last scanned today · Reviewed today
Overview
Unreal MCP Server allows AI assistants like Cursor and Claude Desktop to interact directly with Unreal Engine 5.5+. It bridges the gap between natural language prompts and engine operations by utilizing a native C++ TCP server integrated into the editor alongside a Python MCP wrapper.
Our verdict
This project carries a high risk level with a security score of 25/100, primarily due to the lack of a verified license and the fact that it has not been editorially reviewed. Given the experimental status and the fact that the last commit occurred over 450 days ago, it is not recommended for anything beyond personal testing and local exploration.
- Supports complex actor and blueprint manipulation via natural language.
- Zero direct external dependencies reduces supply-chain surface area.
- Includes a functional sample project for rapid local setup.
- Significant abandonment risk due to last commit 459 days ago.
- Missing license creates ambiguity regarding legal usage terms.
- Experimental status warns of potential breaking changes and instability.
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| ChatGPT | ||||
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf |
Frequently asked questions
›What software and environment versions are required?
This server requires Unreal Engine 5.5 or later and Python 3.12 or higher.
›Is this project suitable for production environments?
No, this project is currently in an experimental state and is not recommended for production use.
›How does the Python server communicate with Unreal Engine?
The Python server connects to the Unreal Engine C++ plugin via a TCP socket on port 55557.
›Can I use this with an existing Unreal project?
Yes, you can copy the UnrealMCP plugin folder into your project's Plugins directory, enable it in the editor, and then build your project solution.
›What is the recommended way to run the Python server?
The server is designed to be run via the UV package manager, as shown in the provided MCP configuration example.
Alternatives
Badge
Maintain this server? Add the live badge to your README.