smart-second-brain

Why we included this project

For Obsidian users who want a chat interface that can actually read the vault, this plugin does something genuinely useful. It embeds your notes into vectors and retrieves the passages most relevant to each question, so answers stay grounded in your own writing rather than in generic model knowledge. RAG responses include reference links back to the source notes, which makes verifying a claim quick and keeps the whole conversation traceable. Privacy is a real design choice here: you can run the entire thing against a local Ollama model with no network calls at all, though OpenAI models remain available for when you want them. Anyone keeping a serious personal knowledge base will find this a practical way to turn a passive note archive into something they can actually interrogate.