Vault-for-LLM
Active·★ 38·MIT·Updated 2026-05-24
★ Trending★ RAG / Knowledge Base★ LLM Infra
🧠 Local-first knowledge system for LLM agents — sqlite-vec + ONNX embeddings, no cloud/Docker/PyTorch dependency
Vault-for-LLM is a local-first memory layer for LLM agents. It creates a portable SQLite knowledge vault from Markdown notes, allowing agents to search and retrieve structured memory on demand. It focuses on agent-oriented memory with bounded retrieval and optional embeddings.
#embeddings#knowledge-base#llm#local-first#mcp-server#mcp-servers#onnx#rag
01
Features
01Local-first SQLite storage
02Keyword + optional vector/hybrid search
03Memory layers (L0-L3) for structured context
04Document Map with bounded citations
05MCP server integration for agent access
02
Compatibility
Python
Python CLI
Verified via docs
MCP
MCP Server
Verified via docs
03
Quick start
1
$ python3 -m venv .venv
2
$ source .venv/bin/activate
3
$ pip install vault-for-llm
04
Use cases
↳Persistent project context for AI agents across sessions
↳Debugging and troubleshooting knowledge retrieval
↳Team collaboration via synced knowledge vaults
05
Alternatives
Brave Search MCP★ 86.5k
Allow your AI Agent to search the real-time internet using Brave Search API. Essential for getting up-to-date information.
FunASR★ 16.6k
Industrial-grade speech recognition toolkit: 170x realtime, 50+ languages, speaker diarization, emotion detection, streaming, and OpenAI-compatible API.
Related searches
Comments
Log in to leave a comment
- QQuinn JohnsonMay 15, 2026
ONNX embeddings mean fast local inference without a running model server.
- Dakota RiveraApr 3, 2026
Local-first knowledge store with no cloud dependency — sqlite-vec does the heavy lifting.
- LLogan WilsonMar 12, 2026
Good for building persistent agent memory that survives process restarts.
- HHayden KimMar 1, 2026
Privacy-preserving by design, nothing leaves the machine.