Knowledge base service: RAG, embeddings (BGE-M3), Qdrant vector search. Reusable across projects.

Persistent memory for DeepSeek Harness: per-conversation notes, selective injection, project memory, semantic Vault search (bge-m3, free optional key) and automatic retrieval — agents stop forgetting.

Fold the DSH skill catalog prompt surface: static KV-cache-stable catalog + BM25/bge-m3 hybrid skill_search + autoRoute hints. Rendering-only fold: source.entries stay untouched, the skill tool and /name injection stay host-owned. v0.3.0 adds a semantic r

Fold the DSH tool surface per request + ChainGuard firewall (high-risk block + exfil-chain detection + anti-obfuscation) + BM25/bge-m3 hybrid tools_search. Shrinks schema tokens 80-90% while keeping selection accuracy. v0.2.0 adds a semantic retrieval leg

Local Transformers.js embedder for Kura (bge-m3). Runs ML models in JS with no Python.

Code RAG MCP server: semantic indexing and retrieval of a local codebase (declaration-aware chunking + bge-m3 embeddings + Qdrant hybrid search). Exposes search_code / index_codebase / index_status / index_cancel / set_root MCP tools and a CLI, plus a loc

Open-source embedding generation for the .cv format. BGE-M3 by default; pluggable.

Kura CLI — dev, build, deploy, and index a Kura docs app (one surface over June).

Kura — the knowledgebase for humans and agents. Portable pure-JS vector retrieval engine (f32 brute-force + binary prefilter + rerank).

NodeJS implementation of @Qdrant/fastembed

OpenAI-compatible embedder for ObjectStack — works against OpenAI, 阿里通义 DashScope, 智谱 BigModel, 硅基流动 SiliconFlow, 火山引擎 Doubao, MiniMax, Ollama, and any drop-in OpenAI-shape endpoint.

Standalone Pi Mentis knowledge extension for Pi >= 0.84.0

ev-agent CLI + evsmem memory server. One-command install, then `evagent -setup`.

Get the user home directory with fallback to the system temp directory

React TypeScript implementation of Material Design 3 Ripple component

Ollama embedding provider and OllamaRag convenience class for @mcp-abap-adt/llm-agent.

Long-term memory for DeepSeek Harness backed by Hypatia: a host-side CLI adapter, a plugin-owned SQLite control ledger, same-request recall, and narrow memory tools. Requires the `hypatia` CLI on PATH.

n8n community node for SiliconFlow AI models - chat completions, vision language models, embeddings, and reranking

Official TypeScript SDK for SIE (Search Inference Engine)

Your Obsidian vault is Claude's memory — local semantic search over your notes via MCP, always fresh, never reindex.