UEMCP 通过两层架构将 AI 助手与虚幻引擎连接起来，将 MCP 服务器（Node.js）与 Python 编辑器插件分离，实现虚幻引擎编辑器的远程部署。该实现为常见的 UE Python API 操作提供了优化的包装，将代码生成减少了高达 85%。该仓库包含为 AI 客户端自动设置、全面的开发上下文，以及三个专门的 Claude 代理，以增强 UE 工作流程。与包管理的 MCP 服务器不同，该仓库设计为可克隆并可能分支，以实现最大定制和开发灵活性。

🚀 快速开始（2分钟）

设置脚本会自动：

• ✅ 检查和安装Node.js（如有需要）（通过 Homebrew、apt、yum 或 nvm）
• ✅ 安装依赖并构建服务器
• ✅ 检测和配置AI开发工具（Claude Desktop、Claude Code、Amazon Q、Gemini Code Assist、OpenAI Codex）
• ✅ 建立你的虚幻引擎项目路径
• ✅ 可选地将UEMCP插件安装到你的项目中

脚本将检测您安装的AI工具并提供配置：

• Claude 桌面与 Claude 代码：原生 MCP 支持

UEMCP - Unreal Engine Model Context Protocol

UEMCP bridges AI assistants with Unreal Engine through a two-tier architecture that separates the MCP server (Node.js) from the Python Editor Plugin, enabling remote deployment of Unreal Engine editors. This implementation provides optimized wrappers around common UE Python API operations, reducing code generation by up to 85%. The repository includes automated setup for AI clients, comprehensive development context, and three specialized Claude agents for enhanced UE workflows. Unlike package-managed MCP servers, this repo is designed to be cloned and potentially forked for maximum customization and development flexibility.

🚀 Quick Start (2 minutes)

# Clone and setup
git clone https://github.com/atomantic/UEMCP.git
cd UEMCP
./setup.sh

# Restart Claude Desktop or Claude Code and test:
# "List available UEMCP tools"
# "Organize the actors in the current map into a sensible folder structure and naming convention"

The setup script automatically:

• ✅ Checks and installs Node.js if needed (via Homebrew, apt, yum, or nvm)
• ✅ Installs dependencies and builds the server
• ✅ Detects and configures AI development tools (Claude Desktop, Claude Code, Amazon Q, Gemini Code Assist, OpenAI Codex)
• ✅ Sets up your Unreal Engine project path
• ✅ Optionally installs the UEMCP plugin to your project

The script will detect which AI tools you have installed and offer to configure them:

• Claude Desktop & Claude Code: Native MCP support
• Amazon Q: MCP support via ~/.aws/amazonq/agents/default.json
• Google Gemini (CLI & Code Assist): MCP support via ~/.gemini/settings.json
• OpenAI Codex: Trusted projects via ~/.codex/config.toml
• GitHub Copilot: Usage instructions provided

📝 Windows Users

Recommended: Use WSL (Windows Subsystem for Linux)

# Install WSL if you haven't already
wsl --install

# In WSL/Ubuntu terminal:
git clone https://github.com/atomantic/UEMCP.git
cd UEMCP
./setup.sh

Alternative: Git Bash

• Install Git for Windows which includes Git Bash
• Run ./setup.sh in Git Bash terminal

Note: The setup script will copy (not symlink) the plugin to your UE project on Windows to avoid permission issues.

Advanced options:

# Specify UE project (automatically installs plugin via copy)
./setup.sh --project "/path/to/project.uproject"

# Install with symlink for UEMCP plugin development
./setup.sh --project "/path/to/project.uproject" --symlink

# Non-interactive mode (for CI/CD)
./setup.sh --project "/path/to/project.uproject" --no-interactive