Build DSH knowledge bases from files, folders, URLs, and text, then retrieve bounded evidence through `knowledge_search`.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-09-08
License
AGPL-3.0
01What can it help you accomplish?
Import DSH knowledge sources and retrieve evidence for model use
Managed knowledge bases with lexical, vector, or hybrid retrieval and ordered evidence context
DSH users who need local or provider-backed document RAG
02How to install into DeepSeek Harness
Prerequisites
- Node.js ^22.19.0 || >=24.0.0, pnpm >=10, initialized DeepSeek Harness, and the listed `allowBuilds` permissions.
Installation steps
- 01
Run `dsh plugin --profile <name> add dsh-knowledge`, then restart the Web service and refresh the page.
$ dsh plugin --profile <name> add dsh-knowledge
Verify the integration
- Use Recall Test before allowing the model to use the knowledge base through `knowledge_search`.
03DSH integration and capability boundaries
DSH knowledge-base and RAG plugin
Knowledge-base management and retrieval
Files, directories, URLs, or plain text→Persisted knowledge bases and 14 model tools including `knowledge_search`
Permanent deletion requires host confirmation.
04Who is it for? When not to use it?
Good for
- DSH users who need local or provider-backed document RAG
Not for
- Intel Macs cannot run onnxruntime-backed local embedding and OCR; use an OpenAI-compatible service or Ollama instead.
- API keys are stored in plaintext locally; protect the profile data directory.
05Compatibility, maintenance and safety notes
- Intel Macs cannot run onnxruntime-backed local embedding and OCR; use an OpenAI-compatible service or Ollama instead.
- API keys are stored in plaintext locally; protect the profile data directory.
AGPL-3.0 source; author README claims captured from the fetched main branch
06Frequently asked questions
Can it work without local embedding models?
Yes. The README says it can run with zero vector configuration; keyword retrieval remains available without downloaded models.
How do I validate retrieval before model use?
Use the Recall Test, then let the model use the base through `knowledge_search`.
What happens when a selected knowledge base is empty or invalid?
The author says it matches zero bases and does not silently expand the search to every base.
07Related DSH workflows
weknora
by tencent
Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
honcho
by plastic-labs
Memory library for building stateful agents
mirage
by strukto-ai
The World's First Virtual Terminal for AI Agents
reme
by agentscope-ai
ReMe: Memory Management Kit for Agents - Remember Me, Refine Me.
08Data and sources
dsh-knowledge 把文档导入、解析、分块、检索、证据组织和模型续读整合进 DSH。它既可以零向量配置运行，也可以完全使用本地模型，不要求额外部署独立的知识库服务。
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-09-08. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.