Sandbox
@chongdashu/unreal-mcp

MCP server for Unreal Engine control

unreal-mcp connects MCP clients to Unreal Engine through a Python server and a native Unreal plugin. The server handles MCP requests, forwards them over TCP, and exposes tools for actor management, Blueprint work, graph edits, and editor control. The included sample project lets you try it without starting from scratch.

2,075 stars343 forksC++Updated 1y ago
Who it's for

Builders who want an MCP client to create actors, edit Blueprints, and steer the Unreal Editor with chat.

What it delivers

You can control Unreal Engine from your agent instead of clicking through the editor by hand.

What it does

Actor management

Create, delete, query, and move actors such as cubes, lights, cameras, and spheres.

Blueprint creation

Create Blueprint classes, add components, set properties, compile Blueprints, and spawn Blueprint actors.

Blueprint graph editing

Add event nodes, function calls, variables, and references in Blueprint graphs.

Editor control

Focus the viewport on actors or locations and adjust camera orientation and distance.

MCP client support

Works with MCP clients such as Cursor, Windsurf, and Claude Desktop through a standard JSON config.

README

Model Context Protocol for Unreal Engine

unreal-mcp

License: MIT Unreal Engine Python Status

This project enables AI assistant clients like Cursor, Windsurf and Claude Desktop to control Unreal Engine through natural language using the Model Context Protocol (MCP).

⚠️ Experimental Status

This project is currently in an EXPERIMENTAL state. The API, functionality, and implementation details are subject to significant changes. While we encourage testing and feedback, please be aware that:

  • Breaking changes may occur without notice
  • Features may be incomplete or unstable
  • Documentation may be outdated or missing
  • Production use is not recommended at this time

🌟 Overview

The Unreal MCP integration provides comprehensive tools for controlling Unreal Engine through natural language:

CategoryCapabilities
Actor Management• Create and delete actors (cubes, spheres, lights, cameras, etc.)
• Set actor transforms (position, rotation, scale)
• Query actor properties and find actors by name
• List all actors in the current level
Blueprint Development• Create new Blueprint classes with custom components
• Add and configure components (mesh, camera, light, etc.)
• Set component properties and physics settings
• Compile Blueprints and spawn Blueprint actors
• Create input mappings for player controls
Blueprint Node Graph• Add event nodes (BeginPlay, Tick, etc.)
• Create function call nodes and connect them
• Add variables with custom types and default values
• Create component and self references
• Find and manage nodes in the graph
Editor Control• Focus viewport on specific actors or locations
• Control viewport camera orientation and distance

All these capabilities are accessible through natural language commands via AI assistants, making it easy to automate and control Unreal Engine workflows.

🧩 Components

Sample Project (MCPGameProject) MCPGameProject

  • Based off the Blank Project, but with the UnrealMCP plugin added.

Plugin (UnrealMCP) MCPGameProject/Plugins/UnrealMCP

  • Native TCP server for MCP communication
  • Integrates with Unreal Editor subsystems
  • Implements actor manipulation tools
  • Handles command execution and response handling

Python MCP Server Python/unreal_mcp_server.py

  • Implemented in unreal_mcp_server.py
  • Manages TCP socket connections to the C++ plugin (port 55557)
  • Handles command serialization and response parsing
  • Provides error handling and connection management
  • Loads and registers tool modules from the tools directory
  • Uses the FastMCP library to implement the Model Context Protocol

📂 Directory Structure

  • MCPGameProject/ - Example Unreal project

    • Plugins/UnrealMCP/ - C++ plugin source
      • Source/UnrealMCP/ - Plugin source code
      • UnrealMCP.uplugin - Plugin definition
  • Python/ - Python server and tools

    • tools/ - Tool modules for actor, editor, and blueprint operations
    • scripts/ - Example scripts and demos
  • Docs/ - Comprehensive documentation

🚀 Quick Start Guide

Prerequisites

  • Unreal Engine 5.5+
  • Python 3.12+
  • MCP Client (e.g., Claude Desktop, Cursor, Windsurf)

Sample project

For getting started quickly, feel free to use the starter project in MCPGameProject. This is a UE 5.5 Blank Starter Project with the UnrealMCP.uplugin already configured.

  1. Prepare the project
    • Right-click your .uproject file
    • Generate Visual Studio project files
  2. Build the project (including the plugin)
    • Open solution (.sln)
    • Choose Development Editor as your target.
    • Build

Plugin

Otherwise, if you want to use the plugin in your existing project:

  1. Copy the plugin to your project

    • Copy MCPGameProject/Plugins/UnrealMCP to your project's Plugins folder
  2. Enable the plugin

    • Edit > Plugins
    • Find "UnrealMCP" in Editor category
    • Enable the plugin
    • Restart editor when prompted
  3. Build the plugin

    • Right-click your .uproject file
    • Generate Visual Studio project files
    • Open solution (`.sln)
    • Build with your target platform and output settings

Python Server Setup

See Python/README.md for detailed Python setup instructions, including:

  • Setting up your Python environment
  • Running the MCP server
  • Using direct or server-based connections

Configuring your MCP Client

Use the following JSON for your mcp configuration based on your MCP client.

{
  "mcpServers": {
    "unrealMCP": {
      "command": "uv",
      "args": [
        "--directory",
        "<path/to/the/folder/PYTHON>",
        "run",
        "unreal_mcp_server.py"
      ]
    }
  }
}

An example is found in mcp.json

MCP Configuration Locations

Depending on which MCP client you're using, the configuration file location will differ:

MCP ClientConfiguration File LocationNotes
Claude Desktop~/.config/claude-desktop/mcp.jsonOn Windows: %USERPROFILE%\.config\claude-desktop\mcp.json
Cursor.cursor/mcp.jsonLocated in your project root directory
Windsurf~/.config/windsurf/mcp.jsonOn Windows: %USERPROFILE%\.config\windsurf\mcp.json

Each client uses the same JSON format as shown in the example above. Simply place the configuration in the appropriate location for your MCP client.

License

MIT

Questions

For questions, you can reach me on X/Twitter: @chongdashu

Files in the repo

Repository payload10 top-level entries
  • .cursor
  • .github
  • Docs
  • MCPGameProject
  • Python
  • .clinerules
  • .gitignore
  • .windsurfrules
  • mcp.json
  • README.md

Discussion (0)

Ask about usage, or say what you built with it

Sign in to join the discussion.

No comments yet. Be the first to say what this is good for.

More connectors

High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

43k

Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code

14k
okf-memory/
okf-agent-memory

Git-native persistent memory for AI coding agents. Implements Google OKF v0.2 with sub-300µs in-memory BM25 search, embedded MCP server, and progressive disclosure. Slashes token bloat by 80% with zero external databases or dependencies. Built in pure Go.

547
tirth8205/
code-review-graph

Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.

31k
2akouwu/
reverify

Stop your AI from making things up — it proposes, deterministic tools decide, every claim checked against ground truth with evidence. Grounded facts and context survive resets. Reverse engineering is the proving ground. MCP server + CLI.

1.1k
t8y2/dbxConnectors

20 MB lightweight cross-platform database client for 90+ databases, including MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, SQL Server, and Dameng. Built-in AI, MCP Server, CLI, desktop and Docker. | 轻量级跨平台数据库管理工具,支持 MySQL、PostgreSQL、SQLite、Redis、MongoDB、达梦等 90+ 数据库,提供桌面端、Docker、CLI、内置 AI 助手和 MCP Server。

19k