DeepSeek Harness Plugins
DeepSeek Harness PluginManifest Valid

dsh-kb-manager

Local knowledge base lifecycle manager with hybrid vector/BM25 search and async indexing for DeepSeek Harness.

Developer ToolsSecurity & Policy
2GitHub Stars0ForksUpdated2026-09-02

Install

$ dsh plugin --profile web add .

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-manager is a local knowledge-base plugin for DeepSeek Harness that provides end-to-end document ingestion, CJK-aware chunking, SQLite-native vector (sqlite-vec) and full-text (FTS5) indexing, and hybrid retrieval using Reciprocal Rank Fusion (RRF). It exposes 13 agent tools and a web settings panel with built-in path whitelisting, SSRF guards, and an asynchronous background job engine.

Key Capabilities

  • 13 unified agent tools covering knowledge base creation, document import, hybrid search, inspection, and index rebuilding
  • SQLite-native indexing utilizing sqlite-vec for KNN vector search and FTS5 for BM25 full-text retrieval without external server processes
  • CJK-aware tokenization using native Intl.Segmenter for Chinese, Japanese, and Korean text
  • Hybrid search pipeline combining vector search, BM25, RRF fusion, and explicit rerank state handling
  • Async job processing system with progress polling, crash recovery, and cooperative cancellation
  • Security-focused ingestion with local path whitelists and SSRF guards against private IP ranges and cloud metadata endpoints
  • Web panel for managing knowledge bases, live job tracking, in-KB hybrid search, and two-step deletions

Useful For

  • Providing long-term memory and document retrieval for DeepSeek Harness AI agents
  • Managing project documentation, codebases, and local text corpora with citation tracing
  • Running offline or private local knowledge retrieval without standing up external vector databases

Who It Fits

  • DeepSeek Harness users requiring local, zero-external-service knowledge retrieval
  • Developers building agent workflows with CJK and multi-format document search needs

Documented Limitations

  • ONNX embedded model loading is structural in the initial release; requires mock or compatible external embedding provider
  • PDF and DOCX document parsers are placeholders in the current MVP (markdown, txt, csv, json, and html are supported)
  • Certain planned v1.1 tools (snapshots, export/import, multi-KB search) exist at service layer but are deferred from agent tool registration

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.

No automated security signals have been published yet.

Source & Registry Notes

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

Source repository
xiaoshi7915/dsh-kb-manager
Registry source
GitHub · dsh-plugin topic
Registry classification
Plugin
Source checked
0b5af12 · 2026-09-04

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
2GitHub stars
Forks
0
Open issues
0
Last commit
2026-09-02
Last release
No release detected
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