UE_ToolkitAI
mcp
Basarisiz
Health Gecti
- License — License: MIT
- Description — Repository has a description
- Active repo — Last push 0 days ago
- Community trust — 26 GitHub stars
Code Basarisiz
- exec() — Shell command execution in core/app_manager.py
- exec() — Shell command execution in core/bootstrap/app_bootstrap.py
Permissions Gecti
- Permissions — No dangerous permissions requested
Purpose
This tool is an AI-powered desktop assistant for Unreal Engine developers that integrates with multiple LLMs (Ollama, OpenAI, Claude) and connects directly to the UE editor. It provides features like asset management, configuration tools, and semantic search to streamline game development workflows.
Security Assessment
The overall risk is rated as Medium. The application makes network requests to communicate with remote LLM APIs and a cloud database (Turso/libSQL). It also runs a local web server (Flask) and opens a local network socket (Port 9998) to talk to the Unreal Engine plugin.
Crucially, the automated audit flagged multiple instances of shell command execution (`exec()`) in `core/app_manager.py` and `core/bootstrap/app_bootstrap.py`. While executing shell commands is common for desktop toolchains trying to launch external processes, it is a high-risk vector. If an attacker were to manipulate the inputs to these functions (e.g., via a compromised LLM response or malicious network request), it could lead to arbitrary code execution on the host machine. No hardcoded secrets or unnecessary dangerous OS permissions were detected, but users should be aware that the app routes data to external cloud services.
Quality Assessment
The project is actively maintained, with its most recent push occurring today. It uses the standard, permissive MIT license and has a clear, comprehensive description. However, community trust and adoption are currently very low, with only 26 GitHub stars. The tool is highly complex (tens of thousands of lines of code), heavily relying on a deep tech stack involving PyQt6, Flask, and external AI integrations.
Verdict
Use with caution: While active and properly licensed, the combination of shell execution code and complex external network communications requires reviewing the source code before handling sensitive or proprietary game projects.
This tool is an AI-powered desktop assistant for Unreal Engine developers that integrates with multiple LLMs (Ollama, OpenAI, Claude) and connects directly to the UE editor. It provides features like asset management, configuration tools, and semantic search to streamline game development workflows.
Security Assessment
The overall risk is rated as Medium. The application makes network requests to communicate with remote LLM APIs and a cloud database (Turso/libSQL). It also runs a local web server (Flask) and opens a local network socket (Port 9998) to talk to the Unreal Engine plugin.
Crucially, the automated audit flagged multiple instances of shell command execution (`exec()`) in `core/app_manager.py` and `core/bootstrap/app_bootstrap.py`. While executing shell commands is common for desktop toolchains trying to launch external processes, it is a high-risk vector. If an attacker were to manipulate the inputs to these functions (e.g., via a compromised LLM response or malicious network request), it could lead to arbitrary code execution on the host machine. No hardcoded secrets or unnecessary dangerous OS permissions were detected, but users should be aware that the app routes data to external cloud services.
Quality Assessment
The project is actively maintained, with its most recent push occurring today. It uses the standard, permissive MIT license and has a clear, comprehensive description. However, community trust and adoption are currently very low, with only 26 GitHub stars. The tool is highly complex (tens of thousands of lines of code), heavily relying on a deep tech stack involving PyQt6, Flask, and external AI integrations.
Verdict
Use with caution: While active and properly licensed, the combination of shell execution code and complex external network communications requires reviewing the source code before handling sensitive or proprietary game projects.
AI-powered desktop toolbox for Unreal Engine developers with multi-model LLM, Function Calling, and real-time UE editor communication
README.md
UE Toolkit - AI-Powered Unreal Engine Toolbox
🎯 项目简介
UE Toolkit 是一个面向 Unreal Engine 开发者的桌面工具箱,旨在通过现代化的技术栈提升开发效率。
核心亮点
- 🤖 多模型 LLM 集成:支持 Ollama 本地模型和 OpenAI/Claude/DeepSeek 云端 API
- 🛠️ Function Calling 系统:26+ 业务工具 + 112 个 UE 编辑器操作工具
- 🔌 MCP 协议集成:标准化的工具发现和调用协议
- 🔍 语义检索:基于 sentence-transformers 和 FAISS 的向量检索
- 🔄 实时通信:与 UE 编辑器的双向通信(Socket + HTTP)
- 💎 Freemium 模式:免费基础功能 + 付费高级功能
✨ 功能特性
1. AI 助手模块
- 多模型支持:Ollama(本地)+ OpenAI/Claude(云端)
- 流式响应:SSE 实时输出,ChatGPT 风格交互
- 工具调用:
- 资产搜索与管理
- 配置模板对比
- UE 蓝图提取与分析
- 日志分析
- 站点推荐
- 上下文管理:多轮对话、会话管理、智能记忆
2. 资产管理模块
- 智能搜索:支持拼音搜索、模糊匹配
- 批量操作:拖放添加、批量导入
- 预览系统:自动生成缩略图、实时预览
- 迁移工具:一键迁移资产到 UE 工程
- 分类管理:自定义分类、标签系统
3. 工程管理模块
- 自动发现:扫描本机所有 UE 引擎和工程
- 快速创建:从模板创建新工程(C++/蓝图)
- 分类管理:自定义工程分类
- 快速启动:右键菜单快速打开工程
4. 配置工具模块
- 模板管理:保存和复用常用配置
- 快速应用:一键应用配置到工程
- 版本对比:对比不同版本的配置差异
5. 站点推荐模块
- 精选 UE 资源站点
- 按分类展示(资源、工具、论坛、学习)
- 一键访问
🏗️ 技术架构
技术栈
AI/ML: Ollama, OpenAI API, sentence-transformers, FAISS, MCP
前端: Python 3.9, PyQt6, QSS
后端: Flask, Vercel Serverless
数据库: SQLite (本地), Turso libSQL (云端)
插件: C++, Unreal Engine 5.6/5.7
工具: Git, PyInstaller, pytest, mypy
系统架构
┌─────────────────────────────────────────────────────────┐
│ 桌面客户端 │
│ ┌──────────────┐ Socket 9998 ┌────────────────┐ │
│ │ AI 助手 │◄────────────────►│ UE 插件 │ │
│ │ Python+PyQt6 │ │ C++ + Python │ │
│ └──────┬───────┘ └────────────────┘ │
│ │ HTTPS │
└─────────┼───────────────────────────────────────────────┘
│
▼
┌──────────────────┐ libSQL ┌──────────────┐
│ Web 服务端 │◄──────────────►│ Turso 数据库 │
│ Flask + Vercel │ │ 东京节点 │
└──────────────────┘ └──────────────┘
核心模块
Client/
├── core/ # 核心系统
│ ├── bootstrap/ # 启动引导(6 阶段)
│ ├── security/ # 授权系统
│ ├── config/ # 配置管理
│ ├── services/ # 核心服务
│ └── utils/ # 工具类
├── modules/ # 功能模块
│ ├── ai_assistant/ # AI 助手(16,129 行)
│ ├── asset_manager/ # 资产管理(11,780 行)
│ ├── my_projects/ # 工程管理(3,715 行)
│ ├── config_tool/ # 配置工具(1,278 行)
│ └── site_recommendations/ # 站点推荐(587 行)
├── ui/ # UI 组件
├── resources/ # 资源文件
└── Plugins/ # UE 插件
├── BlueprintExtractor/ # 蓝图提取插件(112 工具)
└── BlueprintAnalyzer/ # 蓝图分析插件(旧版)
🚀 快速开始
环境要求
- Windows 10/11
- Python 3.9+
- (可选)Ollama(用于本地模型)
安装依赖
# 克隆仓库
git clone https://github.com/Awfp1314/UE_ToolkitAI.git
cd UE_ToolkitAI/Client
# 安装依赖(推荐使用国内镜像)
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple/
运行程序
# 开发模式
python main.py
# 打包为可执行文件
python -m PyInstaller packaging/ue_toolkit.spec
配置 AI 助手
使用 Ollama(本地模型)
# 安装 Ollama # 下载:https://ollama.com/ # 拉取模型 ollama pull qwen2.5:7b # 在程序设置中选择 Ollama 模式使用 API(云端模型)
- 在程序设置中选择 API 模式
- 填写 API Key 和 API URL
- 支持 OpenAI、Claude、DeepSeek 等
📁 项目结构
Client/
├── main.py # 程序入口
├── version.py # 版本号
├── requirements.txt # 依赖列表
├── core/ # 核心系统(~20,000 行)
├── modules/ # 功能模块(~33,000 行)
├── ui/ # UI 组件
├── resources/ # 资源文件
│ ├── icons/ # SVG 图标
│ └── styles/ # QSS 样式
├── Plugins/ # UE 插件
├── Docs/ # 文档
└── tests/ # 测试
🎨 功能演示
AI 助手
- 支持多轮对话
- 流式输出
- Markdown 渲染
- 工具调用(资产搜索、配置对比、蓝图分析)
资产管理
- 拼音搜索
- 批量添加
- 自动生成缩略图
- 一键迁移到工程
工程管理
- 自动发现 UE 引擎
- 从模板创建工程
- 分类管理
- 快速启动
🔧 开发指南
代码规范
- 文件编码:UTF-8
- 日志:使用
get_logger(__name__) - 异常处理:必须记录日志
- 类型提示:推荐使用
typing模块
模块开发
每个模块必须实现 IModule 接口:
from core.module_interface import IModule
class MyModule(IModule):
def get_metadata(self) -> ModuleMetadata:
pass
def get_widget(self) -> QWidget:
pass
def initialize(self) -> bool:
pass
def cleanup(self) -> CleanupResult:
pass
测试
# 运行测试
pytest
# 代码覆盖率
pytest --cov=core --cov=modules
📊 项目统计
| 指标 | 数值 |
|---|---|
| 总代码量 | ~100,000 行 |
| 模块数量 | 5 个 |
| 工具数量 | 138 个(26 业务 + 112 UE) |
| 支持的 UE 版本 | 5.4, 5.6, 5.7 |
| 打包体积 | 50-80 MB(优化后) |
🤝 贡献指南
欢迎贡献代码、报告问题或提出建议!
- Fork 本仓库
- 创建特性分支 (
git checkout -b feature/AmazingFeature) - 提交更改 (
git commit -m 'feat: Add some AmazingFeature') - 推送到分支 (
git push origin feature/AmazingFeature) - 提交 Pull Request
Commit 规范
使用 Conventional Commits:
feat: 新增功能
fix: 修复问题
docs: 文档更新
refactor: 重构
chore: 杂项维护
📄 许可证
本项目采用 MIT 许可证 - 详见 LICENSE 文件
🔗 相关链接
- 官网:unrealenginetookit.top
- 文档:查看文档
- 问题反馈:GitHub Issues
📮 联系方式
- GitHub:@Awfp1314
- 官网:unrealenginetookit.top
- 问题反馈:GitHub Issues
🙏 致谢
- Ollama - 本地 LLM 运行时
- PyQt6 - GUI 框架
- sentence-transformers - 文本向量化
- FAISS - 向量检索
- Blueprint Extractor - UE 插件
⭐ 如果这个项目对你有帮助,请给个 Star!
Made with ❤️ and AI
Yorumlar (0)
Yorum birakmak icin giris yap.
Yorum birakSonuc bulunamadi