Ricardo-Kaminski/local-rag
100% local RAG for Obsidian, Zotero, and Claude Code — LightRAG + Ollama + MCP
Platform-specific configuration:
{
"mcpServers": {
"local-rag": {
"command": "npx",
"args": [
"-y",
"local-rag"
]
}
}
}Add the config above to .claude/settings.json under the mcpServers key.
> 100% local RAG for Obsidian, Zotero, and Claude Code — powered by LightRAG + Ollama
[](https://python.org) [](LICENSE) [](https://github.com/HKUDS/LightRAG) [](https://ollama.com) [](https://modelcontextprotocol.io)
A complete, privacy-first RAG stack that runs entirely on your machine. No API keys. No cloud. No data leaves your computer.
---
| | local-rag | Cloud RAG | |---|---|---| | Privacy | Your data stays on your machine | Sent to third-party servers | | Cost | Free after hardware | Per-token API costs | | Offline | Works without internet | Requires connectivity | | Latency | Local inference | Network round-trips | | Customization | Any model via Ollama | Limited to provider models |
---
┌─────────────────────────────────────────────────────┐
│ YOUR DOCUMENTS │
│ Obsidian vault (.md) │ Zotero storage (.pdf) │
└──────────────┬──────────────────────────────────────┘
│ ingestion pipeline (Python)
▼
┌─────────────────────────────────────────────────────┐
│ LIGHTRAG SERVER │
│ • Knowledge graph (entities + relations) │
│ • Vector index (NanoVectorDB) │
│ • LLM: Ollama → qwen2.5:14b │
│ • Embeddings: Ollama → nomic-embed-text │
│ • REST API at http://localhost:9621 │
└──────────────┬──────────────────────────────────────┘
│
┌───────┴───────┐
▼ ▼
┌─────────────┐ ┌──────────────┐
│ Claude Code │ │ Obsidian │
│ (MCP tool) │ │Smart Connect.│
└───────────Loading reviews...