Skip to content

Repository files navigation

DSphere Internal

License: MIT Node.js

DSphere Internal is a multi-tenant enterprise operations platform: CRM, sales, delivery projects, ticketing, HR, R&D/QA, finance, approvals, and AI-assisted workflows—delivered as a React (Vite) SPA on a Node.js (Express) API backed by MySQL, with optional Feishu/Lark automation, LLM integrations, DSphere CLI, mobile (Expo), and MCP tooling.

Language README
English (this file) README.md
简体中文 README_zh-CN.md
Extended setup & ops guide README-EN.md

Features

  • CRM & sales — Customers, agencies, contacts; quotations, orders, collections, pricing.
  • Projects & delivery — Project lifecycle, deployment assets, schedules, documents, delivery center.
  • Tickets — Ticketing, automation, Feishu group integration.
  • Product & R&D — Product roadmap, requirements, RD projects/tasks, code reviews, iterations.
  • QA — Test cases, inventories, QA jobs (via DSphere CLI).
  • HR & people — HR module, leave, OKRs, recruitment.
  • Finance — Expenses, invoices, counterparties, loans, wallet.
  • Approvals — Approval center and BPM-style flows.
  • AI & agents — LLM gateway, chat/memory, agentic orchestration, skills, MCP exposure.
  • Platform — Multi-tenant settings, roles, audit, subscriptions, licensing UI.

Screenshots

Dashboard

Personalized home with quick module access, task overview, support tickets, calendar, and announcements.

Dashboard

CRM — Customers

Manage customer relationships with filtering, status tracking, location mapping, and sales channel assignments.

Customers CRM

Sales & Finance

Handle quotations, orders, payment collections, and pricing with approval workflows.

Sales Quotes

Executive Dashboard

Real-time KPIs: revenue trends, active clients, prospects, outstanding payments, and geographic distribution.

Executive Dashboard

R&D Workstation

Kanban-style task management with customizable views, iterations, and code review tracking.

R&D Workstation

AI LLM Gateway

Configure multiple LLM providers (DeepSeek, Doubao, Qwen, etc.) with model assignments and access control.

AI LLM Gateway

Agentic Workspace

Create and manage agentic coding tasks linked to DSphere CLI runtimes with browser automation support.

Agentic Workspace

Browser Agent

Technical Center

Manage DSphere CLI runtimes, code agents (Claude Code, Cursor Agent), and automation configurations.

Technical Center

Skill Marketplace

Browse, create, and install AI skills for your CLI environment.

Skill Marketplace

DSphere CLI

Desktop runner for QA automation jobs with multi-platform support (Windows, Mac, Linux).

DSphere CLI

Integrated Email

Built-in email client with Feishu/Lark mail integration.

Email Client

Tech stack

Layer Technologies
Frontend React 19, Vite 5, TypeScript, React Router 7, TanStack Query, ECharts
Backend Node.js (ESM), Express 4, ws
Data MySQL 8, Knex migrations, Prisma 6
Testing Jest (client/server), Playwright (e2e)
Mobile Expo / React Native (mobile/)
Containers Docker Compose (dev, test, release)

Integrations (optional)

Configure via environment and tenant settings as applicable:

  • Feishu / Lark — Bots, OAuth, automation, notifications.
  • Alibaba OSS — Object storage for uploads.
  • LLM providers — OpenAI-compatible adapters (e.g. OpenAI, Anthropic, MiniMax, Tongyi).
  • Email — SMTP / IMAP for mail features.
  • MCP — Model Context Protocol server under server/mcp/ for external AI clients.

Quick start

Prerequisites

  • Node.js 18+ and npm
  • Docker & Docker Compose (recommended for MySQL)
  • MySQL 8 (if not using Docker for the database)

Clone and install

git clone <repository-url>
cd ds-internal
./enable_git_hooks.sh   # macOS/Linux/Git Bash — see README-EN.md for Windows
cd client && npm install && cd ..
cd server && npm install && cd ..

Environment

Create a .env file at the repository root (see README-EN.md for the full template). Minimum database-related variables:

DB_HOST=localhost
DB_USER=root
DB_PASSWORD=password
DB_NAME=dsphere_internal
PORT=8443

Database (Docker recommended)

cd server
docker-compose -f docker-compose-dev.yml up -d
npm run migrate
cd ..

Run (development)

# Terminal 1 — API (from server/)
cd server && npm run dev

# Terminal 2 — web app (from repo root)
npm run dev
  • Frontend dev server: URL printed by Vite (ports from client/vite.config.ts, often 5173).
  • API: http://localhost:8443

Full stack via Docker (release-style preview)

cd server
docker-compose up --build

Open http://localhost:8443.

DSphere CLI (QA runner)

cd dsphere-cli
npm install
cp .env.example .env   # set RUNNER_SERVER_URL, etc.
npm start

See dsphere-cli/README.md for details.

Repository layout

Path Description
client/ React SPA (dsphere-client)
server/ Express API, Knex migrations, Prisma schema
dsphere-cli/ Desktop runner for Playwright QA jobs
mobile/ Expo mobile app
e2e/ End-to-end tests

Documentation

  • README-EN.md — Detailed installation, logging, Docker release scripts, DB/Prisma/test workflows.

Contributing

We welcome issues and pull requests. Please read:

Security

Report vulnerabilities responsibly: SECURITY.md.

License

Copyright (c) 2026 Digital Sphere (Hangzhou) Technology Co., Ltd.

This project is licensed under the MIT License.

Acknowledgments

Built with open-source libraries including React, Vite, Express, Prisma, and many others—see client/package.json, server/package.json, and dsphere-cli/package.json for full dependency lists.

About

DSphere Internal is a multi-tenant enterprise operations platform: CRM, sales, delivery projects, ticketing, HR, R&D/QA, finance, approvals, and AI-assisted workflows—delivered as a React (Vite) SPA on a Node.js (Express) API backed by MySQL, with optional Feishu/Lark automation, LLM integrations, DSphere CLI, mobile (Expo), and MCP tooling.

Resources

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages