An interactive, GPU-accelerated knowledge graph visualization with Graph RAG (Retrieval-Augmented Generation) for Obsidian. Powered by cosmos.gl.
Graph RAG combines vector similarity search with knowledge graph traversal to provide richer, more contextual answers from your vault. Instead of just finding notes that are semantically similar to your question, it also explores the connections between those notes to synthesize comprehensive answers.
- GPU Rendering: Handles massive vaults smoothly via
cosmos.gl - Local Indexing: SQLite-backed vector embeddings
- Contextual Chat: Ask your vault questions and get cited, graph-aware answers
