IDE extensionKnowledge & MemoryVerified Last verified 2026-07-24

Smart Connections (Obsidian)

by Brian Petro

Smart Connections is an open-source Obsidian community plugin that finds related notes and excerpts as you write and lets you chat with your vault. It ships a bundled local embedding model so semantic search works with zero setup and no API key, keeping notes on your device; it can also connect to hosted models if you choose.

Official page

Works with

Obsidian

Community plugins (many AI-powered).

When to use it

Rediscovering related notes you'd forgotten while writing
Asking questions across your own vault instead of manual searching
Keeping note content local rather than sending it to a cloud service

When not to

When you need connections across sources outside your Obsidian vault

What it does

Show semantically related notes and excerpts in a side view while writing
Chat with your notes (Smart Chat) grounded in your vault
Generate embeddings locally with a bundled model — no API key required

How to install

Setup steps with copy-paste config. Sources are cited under each method.

Obsidian community plugins

Obsidian

Requires: Obsidian

Open community plugins
In Obsidian, go to Settings → Community plugins and turn off Restricted mode if needed.
2
Install Smart Connections
Browse for 'Smart Connections', install it, and enable it.
config
```
Settings → Community plugins → Browse → Smart Connections
```
Let it index
On first run it builds local embeddings for your notes; then open the Smart View to see related content.

Smart Connections — documentation · Brian Petro · Documentation · accessed 2026-07-24

How to prompt it

Copy-paste prompts written to exercise this plugin, with what to expect.

Chat with your vault

You want an answer synthesized from your own notes.

Using my notes, summarize what I've written about spaced repetition and list the notes you drew from.

What to expect: Smart Chat retrieves relevant notes via local embeddings and answers with a synthesis plus the source notes, so you can verify it.

Let the initial embedding pass finish so search covers the whole vault.

Tips for getting the most out of it

The bundled local model means you can start with no API key and no data leaving your machine.
Keep the Smart View open while writing to catch related notes in real time.

Prompts to pair with it

Real prompts from the IterPrompt library you can copy or run.

Prompt persona

4.5(147)

research and learn to become top in your field of knowledge

Turn any AI chat into a research and learn to become top in your field of knowledge with this community persona prompt.

1,505 copiesOpen ↗

Prompt persona

4.8(118)

Git Repository Analysis and Knowledge Base Construction

Turn any AI chat into a git repository analysis and knowledge base construction with this community persona prompt.

856 copiesOpen ↗

Skill skill

4.8(126)

Rag Implementation

Build Retrieval-Augmented Generation (RAG) systems for LLM applications with vector databases and semantic search. Use when implementing knowledge-grounded AI, building document Q&

3,724 copiesOpen ↗

Prompt insight-reporting

4.7(53)

Insights Knowledge Base — Product Analytics

Make past analyses findable and reusable instead of lost in slides. Purpose-built for product analytics contexts.

1,145 copiesOpen ↗

Prompt insight-reporting

4.5(114)

Insights Knowledge Base — Marketing Analytics

Make past analyses findable and reusable instead of lost in slides. Purpose-built for marketing analytics contexts.

2,326 copiesOpen ↗

Prompt insight-reporting

4.5(90)

Insights Knowledge Base — Sales Ops

Make past analyses findable and reusable instead of lost in slides. Purpose-built for sales ops contexts.

3,982 copiesOpen ↗

See more prompts in the library

Frequently asked questions

Does it send my notes to the cloud?+

By default it uses a bundled local embedding model, so notes stay on your device. You can opt into hosted models, which would send content to that provider.