A RAG-powered knowledge engine that indexes your local documents and lets AI agents query them. Index your knowledge. Query the cave.
Coming soon — join the waitlist · hello@llmdorado.com
The Scroll
The scroll doesn't just search — it descends. Each level goes deeper into your knowledge graph.
Surface-level recall. Fast retrieval of exact matches from your indexed knowledge base.
Semantic reasoning. The scroll connects related concepts across your documents.
Gap detection. Identifies missing context and proposes where to dig deeper.
Synthesis mode. Generates new insights by combining knowledge from multiple sources.
Output generation. Formats and delivers knowledge as actionable, shareable artifacts.
Built for builders
Your documents never leave your machine. Index Obsidian vaults, PDFs, codebases — anything.
Any AI client can call the scroll via Model Context Protocol. Claude, Cursor, your agents.
Powered by LightRAG + Ollama. Hybrid vector + graph retrieval for deep knowledge traversal.
Built for autonomous agents. The Scroll is how your AI fleet gets context about your world.
LLM Dorado is in active development. Get early access and be the first to query the cave.
Get Early Access→