Plugins

Wissensdatenbanksystem im Cherry-Studio-Stil für DeepSeek Harness (DSH): Datenbanken, Dokumente, Chunking, Embeddings (OpenAI-kompatibel / Ollama / lokal / lexikalischer Fallback), Abruf, modellorientierte Tools und ein Browser-Verwaltungsbereich.

Feishu-(Lark-)Integration für DeepSeek Harness: 25 Tools für Chat-Nachrichten und -Verlauf, Cloud-Dokumente (einschließlich Wissensdatenbank-Seiten, generierter Diagramme und Bilder), bitable-Datensätze, Tabellenkalkulationsbereiche und Drive-Berechtigungen — mit Ihrer eigenen selbst erstellten Enterprise-Feishu-App.

Lokaler Wissensdatenbank-Manager für DSH-Agenten: Multi-Format-Import, CJK-bewusstes Chunking, sqlite-vec-Vektor- + BM25-Hybridsuche mit bge-reranker-base-Rerank, Snapshots, kbpack-Export/Import und KB-übergreifende Suche — 22 Agent-Tools sowie ein Web-Panel.

Dreischichtiges lokales Gedächtnis für DeepSeek Harness: L1 Snapshot-Gedächtnis (MEMORY.md/USER.md), bedarfsgesteuert über die memory_*-Tools gelesen + L2 durchsuchbare Wissensbasis (Fakten + SQLite FTS5, chinesische Suche per Sliding-Window-Tokenisierung) + L3 Import aus Hermes/Claude Code/Codex/WorkBuddy und aus freigegebenen Agent-Hub-Dokumenten (memory/global, memory/projects, akzeptierte Entscheidungen) mit Dedup per Content-Hash. v0.2.0 entfernte jede automatische Injektion: Gedächtnis gelangt nur in den Kontext, wenn das Modell eines der 12 memory_*-Tools aufruft, und Daten bleiben auf dem lokalen Laufwerk.

Offline-Spiegelung der offiziellen DeepSeek-Harness-Dokumentation als dsh-Skill: 168 Website-Seiten (EN/zh) plus 52 nur im Repository vorhandene Referenzdokumente, mit Themennavigation und Volltextsuchindex für die Plugin-Entwicklung.

Knowledge base with a corrective acquisition loop for DeepSeek Harness: when the agent keeps hitting the same failure, a background worker fetches the web under a rate limit, distills the result into a staging area, and only a two-stage commit makes it recallable — pages without sources are refused.

Combines fault knowledge-base symptom-to-cause rules with similar past cases to rank candidate motor diagnoses with evidence terms and suggested extra tests.

Git-native, auditable long-term memory for AI agents

Brings the codebase knowledge-base capability of langchain-ai/openwiki into DSH — one-click generate/read/update of the repository Wiki, Grounded Claims (source-traced knowledge cards) and a project knowledge base built from uploaded local files, reusing the model DSH already has configured, with no second API key.

A living memory workbench for DeepSeek Harness: watch the Hindsight bank grow (7-day curve, recall heat, timeline) and manage knowledge pages and background tasks, with the local daemon auto-started by the plugin.

DSH plugin: auto-upload conversation progress to Tencent IMA (daily note + Work knowledge base), mirroring the local Claude Code ima-upload workflow

Bringt deine bestehende Claude-/ChatGPT-Historie nach DSH: sofortiger Import ohne Tokens der memories.json in einem Claude-Export, lokale Claude-Code-Transkripte ohne Export und Gespräche, destilliert durch das bereits in DSH konfigurierte Modell. Jede Erinnerung trägt wörtliche, vom Code gegen die Quelle geprüfte Belege — eine Umschreibung wird verworfen und gezählt, und nur in Assistenten-Turns gefundene Belege gehen immer in die Prüfung statt in den Speicher.

Legal knowledge-base workbench: hybrid keyword/statistical/vector retrieval over local statute .doc/.docx corpora, faceted filtering, charge-name and civil-cause trees, a document editor, and four AI tools (citation check, contract review, AI analysis, AI drafting).

Code graph and knowledge base over an embedded FalkorDB: AST symbols, caller/callee tracing, refactor blast radius, affected-test discovery, clone detection, cross-language porting contracts and spec-to-code traceability.

loci second-brain sidebar tab: hybrid search, cited Q&A, day-grouped memory browsing and quick capture over a local loci knowledge base; proxies `loci serve-http` same-origin from dsh-better-sidebar, no core fork.

Create and configure Agent presets from a single settings section — persona prompt, a checked knowledge-base entry list injected into the system prompt, individual switches for four knowledge-base tools, per-Agent narrowing of MCP connections, and batch knowledge-base import.

On-demand memory for DeepSeek Harness: session logs are distilled into layered markdown (a summary, an index and per-topic files), and a mem_query tool returns only the matched entries rather than the whole index. Nothing is injected per turn, there are no dependencies, and every entry keeps a source pointer back to the conversation it came from.

几何论（共扼谱几何 CSG）知识库插件 for DeepSeek Harness：以 Markdown 文章为唯一真源，程序自动重建 BM25 索引、类型化主张依赖图与修复工作单并做一致性审计，附 869 条真理层与 SymPy 公式验证。开箱即用的离线知识库（31 个 geo_* 工具），零运行时依赖。Conjugate Spectral Geometry knowledge base plugin for DeepSeek Harness: offline BM25 retrieval, typed

Cross-session knowledge for DeepSeek Harness: ctx.knowledge service + knowledge_* tools, SQLite FTS5 trigram retrieval (CJK substring) with L1 query expansion. V1.11 contract (write/update/search/list/delete + budget + events + error codes).

Knowledge-base browser panel: tree browsing plus drag-to-cite (path-only or full content) into the composer.

Obsidian knowledge-base ingest workflow as a DeepSeek Harness plugin: a bundled skill provider + an obsidian_validate_vault model tool, all declarative and composable.

Offline BM25 knowledge base for the Conjugate Spectral Geometry (CSG) corpus: 208 articles / 3833 chunks / 871 verified truths, 5 geo_* tools (list / search / read / calc / truth) with LaTeX→Unicode rendering, truth→article citation chains and single-character query support, zero runtime deps, `dsh plugin add geometry-knowledge`.

RAGFlow knowledge-base retrieval plugin for the DeepSeek Harness.

The claude-obsidian-derived knowledge-suite skills (wiki, wiki-ingest, wiki-query, wiki-lint, save) as a DeepSeek Harness plugin, with attribution.