🐳 dsh-knowledge-base

A general-purpose knowledge base plugin for DeepSeek Harness (DSH)

Import documents · organize in folders · full-text search · manage in the Web UI

English · 简体中文 · Español · 日本語

---

dsh-knowledge-base turns documents into a searchable knowledge base for DeepSeek Harness agents. Drop in a PDF, Word, or Markdown file — it is parsed, chunked by section, and instantly searchable by your agent with FTS5 (BM25) ranked retrieval. A built-in Web UI lets you manage the knowledge base like a file manager: create/rename/delete folders, move files, and browse content.

✨ Features

• 📥 Import anything — md / txt / json / yml / docx / pdf, no size limit; PDF is parsed in-process with pdfjs-dist (cross-platform, zero system dependencies)
• 🔪 Auto-chunking — documents are split by headings/paragraphs; adjacent small chunks are merged to reduce fragmentation
• 🔁 Upsert — re-importing the same file overwrites it; no duplicate accumulation
• 🗂 Folder management — categories are folders: create / rename / delete (empty) / move files, like an OS file manager
• 🔍 FTS5 full-text search — SQLite FTS5 (trigram tokenizer, no Chinese segmentation needed) + BM25 relevance ranking; falls back to LIKE for short queries or anomalies
• 🤖 Agent-native tools — kb_query / kb_import / kb_list / kb_update / kb_delete, usable directly by the model in conversation
• 🖥 Web management UI — a "Knowledge Base" tab in the conversation view: drag-and-drop import, directory browsing, category management, search
• ⚙️ Configurable categories — no preset categories by default (fits any domain); create folders at runtime from the UI

🚀 Install

dsh plugin --profile web add dsh-knowledge-base

Dependencies: DeepSeek Harness (dsh) provides the @deepseek-ai/* runtime; Node ≥ 22.5 (built-in node:sqlite).

🏃 Quick Start

Via the Web UI (recommended)

1. Start dsh Web: dsh web (or dsh --profile web)
2. Open your browser, create a session, and switch to the Knowledge Base tab at the top of the session view
3. Drag and drop files to import → auto-chunked → browse / rename / move in the directory view
4. Ask your agent in chat: "Use kb_query to search for ISO9001"

Via agent tools (headless / any profile)

import:  Use kb_import to import /path/to/manual.pdf, category "Documents", tags ["manual"]
search:  Use kb_query to search for "transformer"
list:    Use kb_list / kb_list category=Documents
update:  Use kb_update to change id=3's category to "Documents"
delete:  Use kb_delete to delete id=3

⚙️ Configuration

# In the profile's cordis.patch.yml or when installing the bundle
# No categories are preset (fits any domain); entries without a category go to "Uncategorized".
# You can also create folders directly in the UI (persisted, no config edit needed).
- id: knowledge-base
  name: 'dsh-knowledge-base'
  config:
    categories:            # configure as needed, e.g.:
      - Documents
      - Manuals

🗂 Web API (for the UI and third-party integrations)

…