Jingyun DSH Client

Empowering every AI entrepreneur with their own commercial loop

A one-stop AI commercialization desktop client built on Jingyun Studio + DeepSeek Harness (DSH)

Website · Development Docs · Join Community

📖 What is this project?

Jingyun DSH Client is more than just an AI chat client.

It is the deeply integrated product of Jingyun Studio and DeepSeek Harness (DSH) — a complete commercialization platform client that transforms AI agents / skills / workflows into tradable products.

Jingyun injects a complete commercial loop into DSH: membership → subscription & payment → compute governance → asset distribution → private-domain operations, enabling AI developers to package their agents into standalone commercial products within 30 minutes.

🎯 Core Capabilities

💡 Two Ways to Use

# 安装 UI 品牌定制插件
npm install @jingyun-ai/jingyun-dsh

# 在 DSH 配置中启用

git clone <repo-url>
pnpm install
pnpm build
run_pack.bat  # GUI 打包控制台

🏗️ Project Architecture

jingyun_dsh/
├── packages/
│   └── jingyun-dsh/              # [核心] 井云 UI 品牌定制插件 (TypeScript)
│       ├── src/
│       │   ├── client/           # 前端 UI 组件 (React/JSX)
│       │   │   ├── components/   # 通用组件 (登录/设置/会员状态)
│       │   │   ├── modals/       # 弹窗组件 (订阅/充值/支付)
│       │   │   ├── pages/        # 功能页面 (产物面板/自动化/连接器/插件市场)
│       │   │   └── styles.ts     # 全局样式注入
│       │   ├── agent/            # AI 智能体包管理器 (后端)
│       │   ├── routes/           # HTTP 路由 (插件市场/智能体 API/会员接口)
│       │   ├── config/           # 配置加载器 (井云平台对接)
│       │   └── common/           # 公共工具函数
│       └── resources/            # 静态资源 (插件注册表快照、builtin-skills 内置技能包)
├── src-tauri/                    # [桌面] Tauri v2 原生壳 (Rust)
│       ├── src/lib.rs            # 主进程逻辑 (后端进程管理/托盘/窗口)
│       ├── resources/
│       │   ├── splash/           # 启动屏 (HTML)
│       │   └── vendor/           # 内嵌运行时 (Node.js/Python/Git)
│       └── capabilities/         # Tauri 安全权限声明
├── scripts/                      # 构建 & 打包脚本
│   ├── pack_gui.py               # GUI 打包控制台 (Python/Tkinter)
│   ├── build_deps.js             # 依赖构建脚本
│   └── prepare_vendor.js         # Vendor 运行时打包脚本
└── run_pack.bat                  # 一键打包入口 (Windows)

🛠️ Tech Stack

🚀 Quick Start

Requirements

• Node.js >= 18.x
• pnpm >= 8.x
• Rust (stable) — for compiling the Tauri native shell (only required when building the desktop client)
• Python >= 3.8 — only required for the GUI packaging console (optional)

📦 Bundled Environment Packages: Building the desktop client requires embedded runtime packages (Node.js / Python / portable Git), which are too large for the Git repo. Please join the community to get them for free, then place them under src-tauri/resources/vendor/.

Option 1: Install via npm Package (Recommended)

If you already have a running DSH instance, just install the branding plugin:

npm install @jingyun-ai/jingyun-dsh

Then copy and rename packages/jingyun-dsh/desktop-config.example.json to packages/jingyun-dsh/desktop-config.json locally, and configure your Jingyun platform information to integrate the full membership system and commercial capabilities.

Full repo deployment

Install Dependencies

git clone <repo-url>
cd jingyun_dsh
pnpm install

Development Mode (Web)

pnpm build    # 构建插件
pnpm start    # 启动 DSH Web 服务

Open http://localhost:3080 in your browser to get started.

Development Mode (Desktop Client)

pnpm build       # 构建插件
pnpm tauri:dev   # 启动 Tauri 开发模式

Production Build

# 方式一：命令行打包
pnpm tauri:build

# 方式二：GUI 打包控制台（推荐，支持自定义品牌名/域名/租户ID等）
run_pack.bat

📋 Branding

Before local development or packaging, copy and rename packages/jingyun-dsh/desktop-config.example.json to packages/jingyun-dsh/desktop-config.json yourself. Edit packages/jingyun-dsh/desktop-config.json to customize brand and platform integration settings:

{
  "domain": "http://your-domain/",
  "custom_name": "Your App Name",
  "custom_logo": "https://your-logo-url.png"
}

💡 Free Online Activation: Visit jingyun.studio to register an account and get your dedicated domain, with no need to set up your own backend.

You can also configure the application name and domain interactively via the GUI packaging console (run_pack.bat) during the build process.

📦 Plugin Marketplace

This project ships with a DSH plugin registry snapshot (800+ plugins) from the GitHub open-source community. Users can browse, install, and uninstall community plugins with one click via the built-in marketplace.

⚠️ Disclaimer: Plugin registry data is entirely from third-party open-source communities. This software only provides display and a convenient install channel, with no express or implied warranty on third-party plugin security, compliance, or availability. Users bear all risks and losses from installing or using third-party plugins.

🤖 Built-in Skill Packs

Skill packs under the packages/jingyun-dsh/resources/builtin-skills/ directory enhance the AI assistant's domain expertise:

📄 License

This project is licensed under the Apache License 2.0.

Third-party licenses:

• DSH (DeepSeek Harness) — governed by its original open-source license
• Built-in Skill Packs (builtin-skills/) — governed by their original open-source licenses
• Community Plugin Registry (registry-snapshot.json) — governed by its original open-source license; data sourced from third-party open-source communities

🔗 Related Links

• 🌐 Website: jingyun.studio
• 📖 Developer Docs: Feishu Knowledge Base
• 🏪 App Marketplace: jingyun.studio/apps

💬 Join the Community

For questions, suggestions, or to connect with other AI entrepreneurs, scan the QR code to join our WeChat group:

📮 If the QR code has expired, please contact us via the official website for the latest way to join.