Team knowledge base for offline knowledge sharing (FDE box scenario): browse, full-text search and refinement entry points; raw material is queued automatically and distilled into wiki pages by a serial bot session (Karpathy LLM Wiki pattern: immutable raw, two-step refinement, log stream, monthly lint).
Install
# from npm (prebuilt)
dsh plugin --profile web add @weibaohui/dsh-kb
# from GitHub (first run asks for allowBuilds approval — follow the hint, retry)
dsh plugin --profile web add github:weibaohui/dsh-kb
Any plugin you install runs third-party code with your own permissions — it can read your files, use your credentials, and reach the network, and tool approvals don’t sandbox it. GitHub-sourced plugins also run build scripts at install time — pnpm blocks those until you allow them, so an install can stop with ERR_PNPM_GIT_DEP_PREPARE_NOT_ALLOWED or ERR_PNPM_IGNORED_BUILDS; dsh prints the exact key to add under allowBuilds in your profile’s pnpm-workspace.yaml, and the install works on the next run. Allowing a build is a trust decision: only install sources you trust, and pin a commit (github:owner/repo#sha).
README
This plugin publishes its README in Chinese only.
dsh 插件 · 团队知识库：跨会话、跨人的知识共享与沉淀。侧栏入口进全页知识库——目录树浏览、全文搜索、markdown 阅读、raw 素材上传、任意条目 @ 给 agent。上传即自动蒸馏（v0.2+）：素材入队后由 kb-bot 会话按 schema 串行加工成文，也可继续手动 @ 给 agent 精加工。
基于 Karpathy LLM Wiki 模式：raw（不可变素材）→ wiki（成文知识）→ schema（约定），人丢素材、agent 两步加工（分析 → 成文）维护 index 与 log。
v0.2 自动蒸馏：上传即自动入队，bot 会话按 schema 串行加工成文——「⚗️ 蒸馏队列」看进度，蒸馏模型可在设置中指定：
核心功能
- 全页知识库：侧栏「📚 知识库」进入（工艺库下方）；左栏目录树（wiki / raw 懒加载），右栏阅读或搜索
- 多知识库（v0.5.0+）：左栏顶部下拉框切换库（v0.6.3+ 仅显示库名），旁边「＋」弹出小框新建——素材库（建骨架、可上传、自动蒸馏）或产出库（只读浏览/搜索/@），确定后自动选中新库；选中非主库时旁边出现「✕」可移除（不删数据）；对话里「📚 知识库」按钮同样按多库列出
- 库约定可编辑（v0.6.0+）：每个知识库各有一份
schema.md加工约定，顶栏「📐 约定」或 schema.md 阅读页「✏️ 编辑约定」直接改——每库独立，各有特色（页面规范、目录用途、加工流程都可按库定制）；kb-bot 自动蒸馏与交互 agent 加工前都以当前库的 schema.md 为最终权威，保存即对下次加工生效，支持 ⌘S / 恢复默认模板 - 自动蒸馏（v0.2+）：raw/ 新素材自动入队（上传钩子 + 文件监视 + 周期兜底扫描），bot 会话逐个蒸馏；右上角「⚗️ 蒸馏队列」（全局入口，v0.6.2+）看进度/重试/取消/跳转会话，支持暂停、失败退避重试、重启自动恢复、大文件按章节分批成文
- 全文搜索：内置零依赖扫描（大小写不敏感，文本类文件，命中高亮），无需安装任何外部工具
- 问 AI（v0.7.0+）：搜索框输入问题点「🤖 问 AI」——自动检索当前库，把命中页 @ 引用连同问题写入输入框，agent 带着依据作答并附来源
- 版本历史（v0.7.0+）：宿主监视 wiki/ 自动快照（内容有变化才记版本，每页保留 50 份）；阅读页「🕐 历史」查看/恢复任意版本，agent 或人的改动都可回滚
- 页面反馈闭环（v0.7.0+，v0.7.1 起自动执行）：阅读页「⚠️ 反馈」记录问题到本库
wiki/meta/feedback.md；默认提交后后台 agent 自动处理——无人值守会话核实反馈、按 schema 修正页面、把该行标[done]（进度看蒸馏队列面板「⚠️ 反馈处理」区块，可跳转会话）；执行器不可用时自动降级为把指令写入输入框 - 陈旧提示（v0.7.0+）：超过 90 天未更新的页面显示「⏳ N 天未更新」标记，救火前先核对
- 下载导出（v0.7.0+）：任意文档可下载单页原文（markdown/文本/附件）
- markdown 阅读：frontmatter 徽章（标题/作者/标签/更新时间）、
[[wikilink]]、内部链接、图片内联、代码块、表格 - 素材上传：raw/ 区可上传（log/截图/抓包…），服务端强制只进 raw/、拒绝覆盖（raw 不可变）
- @ 给 agent：任意条目一键把
@绝对路径注入输入框，让 agent 阅读或加工 - 骨架自举：首次启动自动创建目录与约定文件（不覆盖已有），零配置
知识库布局
~/.dsh/kb/ # DSH_KB_ROOT 可覆盖（如 /var/lib/dsh/kb）
├── raw/ # 原始素材，不可变（人上传，agent 只读）
├── wiki/
│ ├── howtos/ # 症状 → 原因 → 步骤 → 验证
│ ├── decisions/ # 背景 → 选项 → 结论 → 后果
│ ├── postmortems/ # 时间线 → 根因 → 改进项
│ └── notes/ # 读书摘记/成书笔记（自动蒸馏大文件落这里）
├── index.md # 目录（agent 维护，每篇成文必须挂入）
├── log.md # 操作流水（谁、何时、动了哪些篇）
└── schema.md # 本库加工约定（可编辑，每库独立：页面规范 / 目录用途 / 加工流程）
安装
dsh plugin --profile web add @weibaohui/dsh-kb -w
装完重启 dsh web 即生效。
配套 agent 技能（必须装，否则 agent 不会加工沉淀）
把 skill/dsh-kb/ 安装进技能市场（skills-management），agent 即获得查询/两步加工/矛盾标注/log 纪律。另见 docs/lint-prompt.md：在 dsh-tasks 加一条每月 lint 定时项（死链/孤儿页/陈旧条目清单）。
使用
- 侧栏点「📚 知识库」；顶部搜索框全文检索（Enter），左侧树点开浏览
- 点文件名阅读；
[[链接]]与内部链接可点击跳转；点「@ 给 agent」把条目丢进对话框 - 往 raw/ 传素材：raw 树内打开任意页面，右上「上传素材」
- 加工沉淀：
@raw/xxx.log给 agent，说「按 schema 加工入库」——agent 两步加工后成文、挂 index、记 log - 定制约定：顶栏「📐 约定」编辑当前库的
schema.md（每库独立，保存即生效）——比如某个库只收「解决方案」形态、某个库要求中文标题，各库各具特色
与 dsh-file-share 的分界
file-share 管会话工作区的文件（跟着会话 cwd 走）；dsh-kb 管跨会话、跨人的固定知识库（只读为主 + raw 上传 + 搜索）。二者互不依赖。
联系我 :飞书群
版本兼容性
本插件与 DeepSeek Harness（@deepseek-ai/dsh）的版本对应关系：
| 插件版本 | 适配 dsh 版本 | 备注 |
|---|---|---|
| 0.7.15 | 0.1.7-rc.2 | 当前版本；修复降级路径 Config 导出 null——宿主 settings schema() 不排除 null，会炸 settings.describe() 拖垮整份设置文档（同 dsh-continue#4） |
| 0.7.12 | 0.1.7-rc.2 | 去除 FDE 专用表述，定位为通用团队知识库 |
| 0.7.11 | 0.1.7-rc.2 | 适配 0.1.7 settings 模型（导出 volatile Config，ctx.settings.update 持久化），面板改动重启不再丢失 |
| 0.7.10 | 0.1.7-rc.2 | 当前版本，已在 @deepseek-ai/dsh@0.1.7-rc.2 下验证运行 |
| 0.7.9 | 0.1.7-rc.2 | 已在 @deepseek-ai/dsh@0.1.7-rc.2 下验证运行 |
| 0.7.8 | 0.1.7-rc.2 | 已在 @deepseek-ai/dsh@0.1.7-rc.2 下验证运行 |
发版约定：每次发布新版本时，请在上表追加一行，记录该插件版本实际验证所用的
@deepseek-ai/dsh版本。package.json的engines.dsh声明最低支持版本；本表记录实际验证版本，二者配合使用。
Links
More in this category
tt-a1i/archify#integrations/deepseek-harness★ 92724
Generate validated, self-contained interactive architecture, workflow, sequence, data-flow, and lifecycle diagrams from repositories or system descriptions.
dream-num/dsh-univer-office★ 657
Give DeepSeek Harness a real office environment. Univer Office Plugin brings spreadsheets, docs, slides, canvases, relational tables, and more into one runtime — with connected data, validation, versioned changes, and isolated worktrees for multi-agent collaboration.
ECNU/EduWork#dsh-knowledge-studio★ 275
A workspace-based Studio for source-grounded reports, slides, tables, mind maps, quizzes, flashcards, audio and video, using DSH models and shared artifact services. Office generation requires Python 3.12+ with its dependencies and DSH_OFFICE_PYTHON configured; media rendering requires Chromium, and local ASR additionally requires whisper.cpp and a model.
PerryLink/dsh-industry-research★ 215
Deterministic industry research reports for DeepSeek Harness — company and industry research flows produce structured, verifiable reports from staged evidence.
HuanLinOTO/dsh-plugin-mineru★ 51
Expose MineRU document parsing tools to the model.
PolinniZhong/dsh-knit★ 48
Puts the Markdown documents, source files, images and video that already exist anywhere in the session workspace into the DSH sidebar, ranked by relevance to the current conversation. The default lexical mode matches your recent messages against title, summary and body locally with IDF weighting (BM25) — no model calls, no network, nothing leaves the machine. The list comes from scanning the workspace rather than from a recent-files list, so restarting DSH or opening a new session does not empty it. Results are organised into a three-tier reading order (primary / supporting / related) with a verifiable reason on every row — which field matched, which document cites this one — and one click opens an in-place preview: Markdown as rendered text, relative-path images actually drawn, source code with the host's own syntax highlighting, video streamed over HTTP Range. The same Context Pack is exposed to the agent as a knit_docs tool. Later versions added read evidence and lifecycle tracking (attribution is frozen at the moment of reading), pin / exclude (exclusion applies before retrieval, and is undoable), relation evidence (references / imports / tests, with line numbers) and an incremental index reused across restarts. Since v0.23 there is also an optional hybrid mode: it is off by default and only runs once you configure an Ollama or OpenAI-compatible embedding provider yourself. When it runs, what leaves the machine is your query text plus each candidate file's title and summary — a Markdown summary is an excerpt of the body's first paragraph (at most 60 characters), a code file's summary is its workspace-relative directory, and an image or video contributes its file name; full bodies and absolute paths are not sent, and a loopback provider keeps everything on this machine. Since v0.24 it also associates uncommitted changes: a read-only git status (fixed argv, no shell, whitelisted child-process environment) combined with the same reference / import / test / mention evidence shows which files a change is likely to drag along, and it is off by default. The benefit of semantic retrieval still awaits a real-provider run, so no recall improvement is claimed.
Community comments
Comments are public GitHub Discussions. Loading them connects to GitHub and Giscus; a GitHub account is required to post.