The "rag" topic on GitHub groups 49 open-source projects in the DeepSeek Harness (DSH) ecosystem, led by graphify with 117.1k GitHub stars. graphify — Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. Every project here is indexed by DSH Universe with live GitHub data — stars, activity and install status — so you can compare and install directly.
Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.
Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings and DeepSeek Harness support.
Agent Skills for NVIDIA products — install into Claude Code, Codex, and other coding agents to run Physical AI, robotics, simulation, CUDA, and RAG workflows end to end.
AI novel writing software: organizes inspiration, characters, worldbuilding, outlines, chapter writing, review, and revision into a controllable workflow; offers Windows/macOS desktop versions and Ollama and DSH plugin previews.
Local document knowledge base for DeepSeek Harness: library_add/remove/list, hybrid semantic+keyword library_search with diversity re-ranking, relevance filtering and lost-in-the-middle avoidance, citation-aware injection, library_cite_check and library_diagnose — SQLite-backed index via the storage domain, local embedding, zero model downloads.
Give DeepSeek Harness cross-session long-term memory: the AI automatically remembers who you are, your projects and preferences, new sessions come with the background, personal and work memories are automatically kept separate without interfering, zero-config and seamless | Long-term memory for DeepSeek Harness: the AI remembers who you are, your projects and preferences across sessions, keeps personal and work memories separate, zero-config.
Local-first literature knowledge base RAG (DeepSeek Harness plugin): hybrid retrieval of body text and figure captions, pin a fuzzy memory to the exact passage and figure, one-click DOI to the source. Local-first literature RAG for DSH — turn a fuzzy memory into an exact passage/figure, one-click DOI to source.
Learning assistant extension for DeepSeek Harness (dsh): brings DeepTutor tutoring to your agent — deep explanations, self-test questions, learning paths, personal knowledge-base search (RAG), and note archiving.
Source-grounded deep research, writing, and learning plugin for DeepSeek Harness (dsh): steerable checkpoints, mid-run steering, and citations verified against retrieved sources.
DeepSeek Harness local RAG knowledge base plugin: local document chunking, vectorization, semantic retrieval, running fully locally (Ollama + pure frontend/host side). A local-first RAG knowledge base plugin for DeepSeek Harness.
Learn-by-doing knowledge base: CRAG-style corrective retrieval (L1 Markdown wiki + L2 Hindsight). When the agent keeps hitting the same wall, it rate-limits background web fetching for extra material, distills to staged, and only a two-phase commit enters recall. DSH plugin.
Local knowledge base for DeepSeek Harness: zero-dependency SQLite FTS5 index for markdown/text/code/web pages, source-cited retrieval, session experience library, interactive kbquery Web panel + knowledge galaxy.
Kestrel - a research engine that can't cite what a source never said. Citations are admitted only when the quote is mechanically located in the source; corroboration is counted in independent origins, not source count. DeepSeek Harness plugin + standalone library. Zero dependencies.
Long-term memory plugin for the DeepSeek Harness main body: semantic recall, entity graph, status cards, presets, configurable embeddings; the task board is now a separate dsh-livetaskboard.
Local document RAG retrieval MCP server for DeepSeek Harness — agents search your local .md/.txt docs with source citations. 100% local (fastembed + Chroma), no API keys.
Personal intel station: info you scroll past auto-settles into a searchable knowledge base — hybrid semantic + keyword search with sources, Obsidian feedback, zero daemons zero API key | IntelHub: the first zvec-native personal intel station for DeepSeek Harness