Skills

@ybkk/gate-01

01men/DeptAI-UI-DSH · ★ 3

01men

榕器定制轨(01门)· AI 代理与人类协作前台——行业 AI 工作台(IAW)与一线面板/看板:dsh 一键插件化(班组长工作台/执行卡/Skill 点名直调/钉钉桥),上游 01men/ybkk-AIOS

License
MIT
Added
2026-10-30
Maintenance
active
Install kind
github
Manifest details
Manifest status
Declared
Manifest path
package.json#dsh
Manifest checked
2026-10-30T09:37:50.196Z
Lifecycle scripts
None found

Static manifest checks report declarations only; they are not compatibility or security review.

安装

dsh plugin --profile web add 'github:01men/DeptAI-UI-DSH#4d34ffdb42ff57ceda611f200dfbff24d19ad9ba'

GitHub 源码安装:首次需按提示配置 allowBuilds 构建授权后重试

For install and environment setup, plus plugin development tutorials, see DSH community docs ↗

README

Source README · body: en · Read Chinese README

Table of contents
  1. What is this
  2. Capabilities
  3. Quick start
  4. Architecture
  5. Engineering quality
  6. Documentation
  7. Feedback & licensing
  8. License

简体中文 · English · Agent view (llms.txt)

npm license dsh selftest

dsh plugin add github:01men/DeptAI-UI-DSH

What is this

Gate 01(01门) is the collaboration front-office for AI agents and humans — it brings day-to-day department work (boards, tickets, approvals, industry calculations) into the dsh conversation and panel surface: agents do the work in chat, humans see and govern it in the panel.

  • One-command install: ships as a cordis plugin; once installed it mounts at /gate01 — open /gate01/panel/ and the built-in demo board is already there;
  • Front-office stance: it calls the Yuanbingke governance base 01men/ybkkasi over existing contracts (six managed asset types plus IAM / OIDC SSO / RBAC / scene-level authorization) — the front-office never re-implements governance;
  • Clean exit anytime: dsh plugin remove uninstalls cleanly; reinstall restores everything.

Owned by Hangzhou Yuanbingke Technology Co., Ltd. (brand: Yuanbingke / 元冰可). Positioning record: docs/positioning-rq-dsh-panel-project-20261019.md (Chinese).

Capabilities

All shipped — nothing on this list is a plan.

  • 🏢 Department space workbench — boards for R&D / manufacturing / sales / strategy / finance: KPI cards, ticket flow, approval trails, live event streaming (SSE)
  • 🏭 14 industry scenegraph packs + theme engine — construction machinery, robotics, home appliances, NEV, petrochemical, lithium battery, PCB, medical equipment, steel and more; industries plug in, inactive ones stay browsable in read-only preview
  • 🧮 Manufacturing AI toolkit — 13 calculation kernels + 3 AI tools + 100 curated topics (OEE, Cpk, takt/capacity and other front-line tools) in the manufacturing space
  • 💬 Execution cards in chat — four-state task cards bridging conversation and board, so agent progress is visible the moment it happens
  • 🔐 DingTalk device-flow sign-in — the panel initiates authorization; unauthenticated users get an activation code automatically — zero onboarding friction for the front line
  • 📡 Honest degradation — when the governance base is unreachable or the remote host lacks a capability, a banner says so plainly; nothing fails silently

Quick start

dsh plugin add github:01men/DeptAI-UI-DSH
  1. Open http://127.0.0.1:<dsh port>/gate01/panel/ — the built-in demo board works with zero configuration;
  2. Once connected to the governance base, all live data flows through a localhost proxy (zero cross-origin in the browser); sign-in happens on the base side;
  3. npm package @ybkk/gate-01 and local link: / file: channels are also available (four install channels, details in Chinese).

Architecture

  • Lives inside dsh: panel and cards start with the host; /gate01 is a single port, single entry — no separate deployment surface;
  • Six bundled packages: install artifacts are prebuilt dist — no TypeScript loading, no lifecycle scripts, works right after install;
  • Front-office × base: the data plane reaches the base through the host bridge's localhost proxy (zero browser CORS); identity, permissions, metering and audit all belong to the base.

Engineering quality

  • 🧪 1300+ end-to-end self-tests, all green (1367/1367 measured on 2026-10-29; full regression plus manifest linting before every delivery)
  • 📦 Fresh-install invariant — every delivery is verified on a freshly installed dsh via the plugin flow (installation simulation is part of the self-test)
  • 🚫 Nothing executes at install time — prebuilt dist artifacts, no lifecycle scripts, no runtime compilation
  • 🧯 Fail-closed + honest degradation — security defaults to deny, capability gaps are surfaced, every action is audited
  • 🤖 Agent-friendly — llms.txt provides a machine-readable view; the panel REST API (/api/panel/*) and SSE event stream are open to agents on the same contract

Documentation

Document Contents
Product definition Product shape, install invariant, install channels, boundaries (Chinese)
Agent view Machine-readable fact card for LLMs / agents (Chinese + English)
Enterprise deployment Deployment runbook and base integration (Chinese)
Product one-pager Decision-maker one-pager (online: http://mdzx.fun:8801/product.html)
Governance base The provider of assets & identity — a separate project

Feedback & licensing

  • File issues on GitHub — the repo runs a daily automated patrol; feedback gets triaged within 48 hours;
  • Commercial: priced by team + domain assets; manufacturing customers can opt for project-based private delivery. Contact details are on the product one-pager.

License

MIT © Hangzhou Yuanbingke Technology Co., Ltd.(杭州元冰可科技有限公司)