openbkn-ai/bkn-dsh

OpenBKN's DeepSeek Harness Web UI plugin enables business users to safely, accurately and reliably utilize the enterprise's business knowledge network to complete business analysis and decision-making.

Project Overview项目介绍

bkn-dsh is a native DeepSeek Harness plugin maintained by openbkn-ai that injects a governed OpenBKN business knowledge network into DSH conversations. It is a DSH-specific package built around a pinned upstream revision (currently dsh-v0.1.6-alpha.2) and a version-fenced compatibility bridge, and the README explicitly describes it as "an additive DeepSeek Harness plugin." Integration happens through DSH's native plugin manager: the recommended path downloads an OpenBKN-compatible DSH Runtime archive, verifies it against the adjacent .sha256 file, unpacks it, and runs ./openbkn-dsh-runtime-*/bin/dsh web (or bin\\dsh.cmd web on Windows). The runtime keeps its profile under an isolated OPENBKN_DSH_HOME so the user's existing ~/.dsh is left untouched; no pnpm or registry access is required on first start.

In a typical session the user clicks OpenBKN in the DSH Web sidebar, supplies the platform address, completes the guided CLI sign-in or pastes a platform token, tests the connection, picks an authorized business knowledge network, and starts a conversation inside that network's local workspace. The product is aimed at business analysts who need explainable answers, at knowledge-network owners who want their semantics reused consistently, and at enterprise AI platform teams that need least-privilege knowledge access without exposing unrestricted data interfaces. A trial path is documented using bkn-foundry to run a local platform (deploy/dev/mac.sh on macOS, or Docker engine with at least 16 GB of memory) and bkn-samples to import the supply_ontology_hand end-to-end dataset.

Dependencies and constraints are strict: the plugin only supports the currently pinned DSH revision and requires Node.js ^22.19.0 or >=24.0.0; Runtime archives are published only for darwin-arm64 and win32-x64 (no darwin-x64). The compat/dsh-0.1.6-alpha.2/apply.mjs patch is mandatory for source-build trees — without it stored session events are rejected on every DSH restart — and it must be reverted before changing DSH versions. Other caveats include: running DSH directly from source in dev form breaks tool dispatch for any plugin, a cancelled or failed turn between bkn_start_interaction and bkn_finish_interaction leaves an unclosed Interaction on the platform side (automatic closing is future work), OpenBKN tokens must never be placed in Cordis YAML, and an upstream-dsh-watch workflow opens tracking issues when upstream releases move past the pin.

bkn-dsh 是面向 DeepSeek Harness(DSH)的官方原生插件,由 openbkn-ai 维护。它把经过治理的 OpenBKN 业务知识网络注入 DSH 对话中,让授权用户在一次会话里选定一个业务知识网络(businessDomain),在显式且不可变的作用域内分析,并能查看每条回答背后的业务来源(provenance)。推荐安装方式是下载与 DSH 0.1.6-alpha.2 锁定的 OpenBKN Runtime 压缩包,解包后执行 ./openbkn-dsh-runtime-*/bin/dsh web,校验 .sha256 后启动;该 Runtime 仅依赖 Node.js ^22.19.0 或 ≥24.0.0,并提供 darwin-arm64 与 win32-x64 两种构建,不发行 darwin-x64。

典型使用流程:在 DSH Web 侧边栏点击 OpenBKN,填入平台地址并完成 CLI 登录或粘贴平台令牌,测试连通性后选择一个已授权的业务知识网络,为其创建本地工作区并发起会话。该插件面向业务分析师、知识网络所有者以及企业 AI 平台团队,适用于需要可解释、带来源审计且最小权限访问的企业决策场景;trial 路径可结合 bkn-foundry 在本地或 macOS 脚本部署平台,并从 bkn-samples 导入 supply_ontology_hand 作为端到端样本。

依赖与前置条件:仅支持当前钉住的 dsh-v0.1.6-alpha.2 一个上游 DSH 版本;用户在源码构建时必须先应用 compat/dsh-0.1.6-alpha.2/apply.mjs 兼容性补丁,否则插件持久化的事件会在每次重启时被拒绝。已知的限制包括:插件必须在打包形态下运行(直接 dev 启动会破坏所有插件的工具派发)、取消或失败的回合可能留下未关闭的 Interaction、自动收尾属于未来工作,且不得把令牌写入 Cordis YAML;升级 Runtime 时只需解压新包到新目录,配置与会话由 OPENBKN_DSH_HOME 隔离目录跨版本保留,卸载请使用 dsh plugin remove @openbkn/dsh-business-context 并清理残留的 node_modules/@openbkn。

Pre-install check安装前体检Compatibility · Security兼容性 · 安全性 3 warnings3 项注意
  • No license declared - all rights reserved by default; ask the author before commercial use or redistribution未声明开源许可证 —— 默认「保留所有权利」,商用或再分发前先问作者
  • Only 4 stars - very few users, little community feedback星标只有 4,几乎没人在用,遇到问题缺少社区反馈
  • No DSH plugin manifest detected - it may only carry the dsh-plugin topic, so the install method must be confirmed on the spot未检测到 DSH 插件清单:可能只是打了 dsh-plugin 话题,安装方式要现场确认
DSH walks through these 9 checksDSH 会逐条核对这 9 项

Compatibility兼容性

  • DSH, Node, OS and profile requirementsDSH 版本 / Node 版本 / 操作系统 / profile 是否满足要求
  • External dependencies and runtimes (Electron / Python / Docker, ...)外部依赖与运行时(Electron / Python / Docker 等)是否齐备
  • Conflicts with installed plugins: command names, skill / tool names, ports, duplicate MCP registration与已装插件是否冲突:命令名、skill / tool 重名、端口占用、重复 MCP 注册

Security安全性

  • Repo matches the facts registered here; archived or abandoned?仓库是否与页面登记一致,是否归档或长期停更
  • Safety of preinstall / install / postinstall and install.sh / setup.ps1preinstall / install / postinstall 与 install.sh、setup.ps1 是否安全
  • curl|bash, download-then-execute, obfuscation, unrelated domains → stop immediatelycurl|bash、下载即执行、混淆代码、无关域名 → 立刻停止
  • Typosquatting or unmaintained packages among the new dependencies新增依赖里有没有 typosquatting 或无人维护的包
  • Requested permissions vs. what the feature actually needs申请了哪些权限、是否超出功能所需(filesystem / network / shell / clipboard)
  • Any sudo / admin requirement, plus uninstall and rollback是否要求 sudo / 管理员权限,以及卸载与回滚方式

Anything uncertain must be marked unknown with a note on how to confirm it. This site's signal screen is a static snapshot, not a security audit.拿不准的必须标「未知」并说明要我怎么确认。本站的信号筛查是静态快照,不能替代安全审计。

Or use CLI install (for developers)或使用命令行安装(适合开发者)

CLI Install命令行安装

dsh plugin --profile web add @openbkn/dsh-business-context@0.2.0-rc.2-openbkn.0.2.0-2

把 openbkn-ai/bkn-dsh 加入你的 DSH 配置(web profile)即可启用。

READMEREADME

bkn-dsh

中文

Bring governed OpenBKN business knowledge into DeepSeek Harness conversations.

Why

Enterprise decisions need trusted business objects, metrics, rules, relationships, and permission-aware evidence. General-purpose chat alone does not provide that governed context.

What

bkn-dsh is an additive DeepSeek Harness plugin that lets authorized users select one OpenBKN business knowledge network for a conversation, analyze within that explicit scope, and inspect available business provenance for each completed answer.

Version prerequisite: the provenance views need a valid OpenBKN token and a businessDomain the deployment allows (x-business-domain; the stock chart allow-lists bd_public, no admin action needed). Verified against OpenBKN 0.1.4: the observability read routes have no license gate, so community deployments show the full provenance panel; the business graph may carry fewer enriched references there (enterprise optimizer, unverified). No upgrade prompt is ever shown for reads — a 403 always means the domain or account was refused.

Who it serves

  • Business users and analysts who need explainable, governed analysis.
  • Knowledge-network owners who want their business semantics reused consistently.
  • Enterprise AI platform teams that need least-privilege knowledge access without exposing unrestricted data interfaces.

Business value

  • Grounds analysis in authorized business semantics and current platform data.
  • Keeps each conversation in one explicit, immutable knowledge-network scope.
  • Preserves native DeepSeek Harness workflows while adding traceable OpenBKN context.
  • Reduces the cost of reaching trusted business insight without requiring query-language expertise.

The published package README contains the same product overview for package consumers.

Install and start

From plugin 0.2.0-rc.2-openbkn.0.2.0-1 on, an unpatched DeepSeek Harness 0.2.0-rc.2 is all you need: install the plugin package with DSH's own plugin manager. The same package works on all three forms of DSH:

DSH form How you run it
Official desktop app 0.2.0-rc.2 the DeepSeek Harness app
npm CLI npm install -g @deepseek-ai/dsh@0.2.0-rc.2, then dsh web
Source checkout, built dsh-v0.2.0-rc.2 checkout after pnpm install && pnpm run build, then node apps/cli/lib/bin.js web

Showing the opening section of the README — the full document lives in the repository以上为 README 开头摘要,完整文档在仓库内 · View the full README on GitHub →在 GitHub 查看完整 README →

← 上一个 Prev dsh-unknown-theme 下一个 Next DSH-SW-and-CAD →