← DeepSeek Harness Plugins
DeepSeek Harness PluginManifest Valid

dsh-kb-rag

Local-first literature RAG plugin for DeepSeek Harness with exact passage citations and Zotero integration.

VisionUI & ProductivityDeveloper ToolsBrowser & WebSkills & Workflows
15GitHub Stars5ForksUpdated2026-10-21

Install

$ dsh plugin --profile web add [email protected]

Plugin Overview & Capabilities

AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.

source-grounded
dsh-kb-rag is a local literature knowledge base and RAG plugin bundle for DeepSeek Harness. It indexes PDFs, text documents, and Zotero libraries into a local SQLite database, combining BM25 keyword search and BAAI/bge vector embeddings with reranking. It provides 10 agent tools for document ingestion, retrieval, deduplication, and formatted citation reporting without external API calls.

Key Capabilities

  • Hybrid retrieval combining BM25 keyword search, bge-small vector cosine similarity, and bge-reranker reranking
  • Section-aware document chunking with weighted abstracts and methods for PDF, TXT, MD, and DOCX files
  • Passage-level provenance including physical page numbers, DOI links, and in-library Zotero reference cross-referencing
  • Local background ingestion pipeline with deduplication, SHA-256 skip checks, and async job polling
  • Interactive and self-contained HTML answer generation across 10 structured presentation templates

Useful For

  • Querying academic paper libraries directly from DeepSeek Harness with exact page and section citations
  • Syncing and querying local Zotero collections and attached PDF papers
  • Offline, privacy-preserving technical documentation search without cloud API data leakage

Who It Fits

  • Academic researchers and students managing large PDF literature collections
  • DeepSeek Harness users requiring private, local document RAG capabilities
  • Knowledge workers seeking direct passage-to-source attribution for research synthesis

Documented Limitations

  • Scanned PDFs lacking an embedded text layer are skipped; OCR is not supported
  • Cross-language retrieval is limited at the engine level because queries are not automatically translated internally
  • Downloads approximately 1.2 GB of local embedding and reranking models on first use and requires Python dependencies
  • Page-level navigation is limited to PDF files; TXT, MD, and DOCX support only section-level references

DSH Compatibility

Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.

Not tested yetNo runtime compatibility tests have been published yet.

Security Signals

Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.

Dependency Counts

package.json declares 0 runtime, 0 development, 0 peer, and 0 optional dependencies.

info
Dsh Bundle Declared

package.json declares DSH bundle metadata.

info
License Declared

GitHub reports the repository license as MIT.

info
Package Manifest Available

A root package.json was captured and can be inspected by the registry.

info
Source Available

Public GitHub source metadata is available for this registry snapshot.

info

Source & Registry Notes

Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.

Source repository
Breeze136/dsh-kb-rag
Registry source
GitHub · dsh-plugin topic
Registry classification
Plugin Bundle
Source checked
3a0b7e7 · 2026-10-21

This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.

Repository Activity

GitHub stars
15GitHub stars
Forks
5
Open issues
4
Last commit
2026-10-21
Last release
2026-10-21
For maintainers

Maintaining this plugin?

This listing is generated from public repository data. If you maintain this project, you can review the information and share this listing with your users if you find it useful.

Add to README
Listed on DSHPlugin.app