Name: Smart Second Brain – Obsidian Plugin · Obsidian Stats
Rating: 78 (1377 reviews)
Author: Leo310, nicobrauchtgit

by Leo310, nicobrauchtgit

Score: 78/100

Install Code

Description

Category: Learning & Knowledge Management

#ai #knowledge-management #chatbot #local-llm

The Smart Second Brain plugin enhances knowledge management in Obsidian by integrating AI-powered assistance to interact with notes. It enables users to chat with their notes using a retrieval-augmented generation (RAG) pipeline, offering references to original sources within the vault. The plugin supports various large language models, including local models via Ollama and cloud-based models like ChatGPT. Users can switch between models based on their needs and preferences. The plugin operates offline to ensure privacy and provides options to save and resume chat sessions.

Reviews

No reviews yet.

Stats

1257

stars

66,742

downloads

119

forks

937

days

217

total PRs

open PRs

closed PRs

198

merged PRs

237

total issues

open issues

207

closed issues

592

commits

Latest Version

2.1.0

3 days ago

Changelog

Tags are now also respected in the smart graph

Smart graph

Tags as nodes. A new Scope → Tags toggle draws each tag as a node linked to every note that carries it, matching Obsidian's own graph with its Tags filter on. Tag nodes use the theme's tag colour, size by how many notes carry them, and clicking one opens Obsidian's search on that tag. Nested tags are their own nodes. (#473)
Tags shape topics. Tag edges feed topic detection alongside wiki links, in both fused and link-only mode, so a vault organised by tags no longer reads as "nothing is linked". Each tag edge is damped by the tag's breadth (1/√n over the notes carrying it): a topical tag on a dozen notes binds them firmly, while a status tag on hundreds can't drag unrelated notes into one topic. Tags are never topic members: they carry no cluster of their own, don't count toward topic size, and don't fold into collapsed bubbles. (#474)
A topic's own tags sit inside it. Tags that live in a topic (at least half of their notes are members) are drawn within its region, pulled to its centre, and can anchor its label. (#476)
Inferred links need an index, and the switch says so. With no graph embedding index selected (or a selection whose index was removed), the Inferred links and Highlight inferred links toggles are disabled with a hint pointing at the Graph settings. The stored preference is kept, so they come alive as soon as an index is picked. (#475)
Changing the graph index rebuilds the graph. Picking or clearing the index in settings previously left the old inferred edges on screen, still shaping topics, until an unrelated setting forced a rebuild. It now rebuilds on the spot; switching back to a previously used index is served from cache. (#475)

See all version on GitHub

README file from

Github

Your Smart Second Brain is a free and open-source Obsidian plugin that makes your vault smarter: better search, an interactive knowledge graph, and an AI assistant that actually knows your notes.

Search surfaces your notes, the graph groups them into the topics your vault is actually about, and the assistant can draw on both when answering. Search and the graph work right away with no AI provider. Connecting one unlocks the full agent. Runs on desktop and mobile.

smartsecondbrain.dev has a demo video, the features, setup guides, and documentation.

Features

Search that combines keyword matching with meaning, so you find the note you were thinking of rather than the one you typed (how it works)
Smart Graph that works out what your vault is about: notes grouped into topics from your links, your tags, and what you wrote, with nothing to maintain (how it works)
Agents that read and write your notes, with skills, memory, MCP, and integrations with your other plugins

Install

Install Smart Second Brain from Obsidian's community plugins and enable it. Search and the Smart Graph work immediately, with no configuration.

Add an embedding model and pick it for search and for the graph to unlock semantic search and topics that read what your notes say, not just how they link; add an AI provider to enable the agent. See the getting started guide.

Out of the box, nothing leaves your machine. The only network requests are to the providers you configure, and there is no telemetry — see privacy.

Contributing

Report a bug or request a feature here
Ask a question in the Q&A
Share a skill, agent, or workflow in Show and tell
Pull requests welcome. CONTRIBUTING.md has the development setup and how we handle AI-assisted contributions; AGENTS.md has the architecture, commands, and conventions

We initially built this as a university project. That's long finished, but we keep developing it in our spare time, as an experimental playground for state-of-the-art AI and as a tool for the Obsidian workflow we're passionate about. Much of the code is written with AI coding agents; we review, test, and stand behind every change, and commits and pull requests say when an agent was involved.

Smart Second Brain