OMEGA Memory
unlistedby OMEGA
Search your vault by meaning, rediscover forgotten notes, and detect contradictions. Local-first, powered by ONNX embeddings.
OMEGA Memory
Semantic search, knowledge rediscovery, and agent memory for your Obsidian vault. Find notes by meaning, not just keywords.
Features
Semantic Search
Search your vault by meaning. "What storage approach did we decide on" finds your note about "database choice" even though the words don't match. Powered by a local AI embedding model (bge-micro-v2, runs entirely on your machine).
Rediscover
Surfaces notes you haven't opened recently that are related to what you're working on now. Turns your vault from a graveyard into a living knowledge system.
Agent Memory
If you use OMEGA with Claude Code, Cursor, or Windsurf, see your coding agent's accumulated memories inside Obsidian. Decisions, lessons, and context from your coding sessions, all in one place.
Contradiction Detection
Scan your vault for notes that contain conflicting information. Results appear in a modal with similarity scores.
Duplicate Detection
Find near-identical content scattered across different notes. Keep your vault clean.
How It Works
OMEGA Memory indexes your vault using a local AI embedding model. The model runs inside your browser via WebAssembly, so your notes never leave your machine. On first install, the model (~17MB) downloads once and is cached for future use.
- Read-only: The plugin never modifies your vault files
- Local-first: All processing happens on your device
- No API keys: Works without any external accounts or configuration
- Zero config: Install, enable, search
Installation
From Community Plugins (recommended)
- Open Obsidian Settings
- Go to Community Plugins, click Browse
- Search for "OMEGA Memory"
- Click Install, then Enable
Manual
- Download
main.js,manifest.json, andsql-wasm.wasmfrom the latest release - Create a folder
.obsidian/plugins/omega-memory/in your vault - Copy the three files into that folder
- Enable the plugin in Obsidian Settings > Community Plugins
Usage
- Click the search icon in the left ribbon, or use Cmd/Ctrl+P > "OMEGA: Semantic search"
- Type a question or concept in the search box
- Click any result to open the note
- The Rediscover section updates as you navigate between notes
- Use Cmd/Ctrl+P > "OMEGA: Find contradictions" to scan for conflicting content
- Use Cmd/Ctrl+P > "OMEGA: Find duplicates" to find overlapping content
OMEGA Integration
OMEGA Memory works standalone with no dependencies. If you also use OMEGA (pip install omega-memory) for AI agent memory, the plugin automatically detects it and shows your agent's memories in the sidebar.
OMEGA Pro ($19/mo) unlocks:
- Agent Bridge: Your coding agent can search your Obsidian vault during sessions
- Multi-vault: Index and search across multiple vaults
- Cloud sync: Sync your index across machines
Network Usage
This plugin makes the following network requests:
- First use: Downloads an AI embedding model (~17MB) from HuggingFace Hub via cdn.jsdelivr.net. Cached locally after download.
- First use: Downloads sql-wasm.wasm (~644KB) from cdn.jsdelivr.net. Cached locally after download.
- Pro license only: If you enter a Pro license key, it is validated once against admin.omegamax.co. The key is cached for 24 hours.
Your notes are never sent to any external server. All search and indexing happens locally on your machine.
Performance
- First install: ~15-30 seconds (model download + initial indexing)
- Subsequent launches: ~2-5 seconds (model load from cache)
- Search latency: <200ms per query
- Indexing: ~32 notes per second (batched embedding)
- Bundle size: ~91KB
Commands
| Command | Description |
|---|---|
| Semantic search | Open the search sidebar |
| Memory timeline | Chronological view of recent vault activity |
| Re-index vault | Force re-index all notes |
| Find contradictions | Scan vault for conflicting content |
| Find duplicates | Find near-identical content across notes |
License
Copyright 2025-2026 Kokyo Keisho Zaidan Stichting
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.