Embedding

pending

by Markcode

Semantic similarity for notes using embeddings.

3 starsUpdated 2mo agoMITDiscovered via Obsidian Unofficial Plugins
View on GitHub

Embedding (Obsidian plugin)

Semantic, keyboard-first navigation for your vault. This plugin builds embeddings for your notes and opens a floating pop-out panel so you can jump through related ideas without touching the sidebar.

Embedding pop-out panel

Currently supports OpenAI embeddings via an OpenAI API key.

What it feels like

  • Open the pop-out and stay on the keyboard.
  • a always opens the original note.
  • b/c/d… open ranked similar notes.
  • z recomputes using the currently open note as the new “original”.
  • Hotkeys are captured while the panel is open, so focus changes do not break navigation.

Features

  • Floating similarity panel (ESC closes it)
  • Single-letter navigation with deterministic hotkeys
  • Batch embedding updates for all notes
  • Optional auto-update on startup (same logic as manual update)
  • Configurable model, dimensions, and API base URL

MCP server (optional)

Enable the MCP JSON-RPC server in settings to expose semantic search over http://127.0.0.1:<port>/mcp. Tools: semantic_search_text, semantic_search_note, fetch_note.

Commands

  • See connections for current note
  • Update all note vectors

Settings

  • API key
  • API base URL
  • Model
  • Dimensions
  • Max input chars
  • Similarity limit
  • Batch size
  • Auto update on startup

Installation (manual)

  1. Copy this folder to YOUR_VAULT/.obsidian/plugins/embedding/
  2. npm install
  3. npm run build
  4. Enable the plugin in Obsidian

Usage

  1. Configure your API key in settings.
  2. Run Update all note vectors once to generate embeddings.json.
  3. Run See connections for current note to open the pop-out panel.

Data

  • Embeddings are stored in embeddings.json at the vault root.
  • The file is created automatically if missing.

For plugin developers

Search results and similarity scores are powered by semantic analysis of your plugin's README. If your plugin isn't appearing for searches you'd expect, try updating your README to clearly describe your plugin's purpose, features, and use cases.