A RemNote plugin that provides a generic, extensible bridge for external tools to interact with your RemNote knowledge base via local WebSocket APIs. It powers MCP servers, CLI apps, and broader automation flows, including but not limited to AI assistants.
This is a working solution, but still experimental. If you run into any issues, please report them here. Further it is an improved and renamed fork of the original plugin MCP Bridge plugin by Quentin Tousart.
This project is a bridge layer with two consumer paths:
- MCP Server path:
- RemNote Automation Bridge (this project): RemNote plugin exposing RemNote API via WebSocket
- RemNote MCP Server: companion server exposing MCP tools to AI assistants
- Demo: View MCP Server Demo →
- CLI app path (e.g. for OpenClaw):
- This bridge plugin remains the RemNote endpoint
- RemNote CLI: companion app for integrating RemNote with OpenClaw and other agentic workflows via the same WebSocket bridge
- Demo: View RemNote CLI Demo →
For both paths always 2 components are required - the bridge and either the MCP server or the CLI app.
Model Context Protocol is an open standard by Anthropic that allows AI assistants to interact with external tools and data sources. With this plugin, your AI assistant becomes a true PKM companion.
- Create Notes - AI can create new notes with titles, content, and tags
- Search Knowledge Base - Full-text search across all your Rems
- Read Notes - Access note content and hierarchical children
- Update Notes - Modify existing notes, append content, manage tags
- Daily Journal - Append entries to today's daily document
- Sidebar Control Panel - Monitor MCP connection status, statistics, and action history
- Auto-tagging - Automatically tag notes created via MCP (configurable)
- Session Statistics - Track created/updated/journal entries/searches
- Action History - View last 10 MCP actions with timestamps
- Configurable Settings - Customize behavior through RemNote settings
- Real-time Status - Live connection status indicator in sidebar panel
This plugin connects your RemNote knowledge base to AI assistants through a locally running MCP server (installed separately from this plugin). Here's the data flow:
RemNote ↔ Local MCP Server ↔ AI Assistant
- The plugin communicates exclusively with your local MCP server via WebSocket (default:
ws://127.0.0.1:3002) - The MCP server forwards data to your AI assistant (Claude, GPT, etc.) using the MCP protocol
- The plugin itself does NOT send data to external servers - all external communication happens through your local MCP server
Your RemNote data is only shared with the AI assistant you've configured in your MCP server setup. The plugin acts as a local bridge and has no built-in external network access beyond the local WebSocket connection.
For technical details about the security model, see the RemNote MCP Server Security Model documentation.
Currently, this plugin is only available for local development installation. Future releases may be published to RemNote's plugin marketplace.
Development Installation:
-
Clone and build the plugin:
git clone https://github.com/robert7/remnote-mcp-bridge.git cd remnote-mcp-bridge npm install npm run dev -
In RemNote:
- Go to Settings > Plugins
- Click "Develop from localhost"
- Click "Develop"
- The Automation Bridge (MCP, OpenClaw...) plugin will now be active
-
Access the control panel:
- Look for the MCP icon in RemNote's right sidebar toolbar
- Click the icon to open the control panel
Beginner walkthrough with screenshots:
- Run The Plugin Locally (Beginner Guide)
- Execute Bridge Commands from RemNote Developer Console (Screenshot Walkthrough)
Important: The plugin alone is not sufficient - you must also install the RemNote MCP Server, which connects your AI assistant to this plugin.
Install the server globally:
npm install -g remnote-mcp-serverFor detailed installation instructions, configuration, and troubleshooting, see the RemNote MCP Server repository.
Multiple AI agents can connect to the MCP server simultaneously, but the system enforces a single RemNote plugin connection. This means:
- Multiple AI assistants (e.g., multiple Claude Code sessions) can access the same RemNote knowledge base concurrently
- The MCP server uses HTTP Streamable transport, supporting multiple concurrent client sessions
- However, only one RemNote app instance can be connected at a time via the WebSocket bridge
- This is a RemNote plugin limitation, not an MCP server limitation
For technical details about multi-agent support and connection architecture, see the RemNote MCP Server documentation.
Access plugin settings in RemNote via Settings > Plugins > Automation Bridge (MCP, OpenClaw...):
| Setting | Description | Default |
|---|---|---|
| Auto-tag MCP notes | Add a tag to all AI-created notes | true |
| Auto-tag name | Tag name for AI-created notes | MCP |
| Journal entry prefix | Optional prefix for journal entries | `` |
| Add timestamp to journal | Include time in journal entries | true |
| WebSocket server URL | MCP server connection URL | ws://127.0.0.1:3002 |
| Default parent Rem ID | Parent for new notes (empty = root) | `` |
Once connected, your AI assistant can use these tools:
| Tool | Description |
|---|---|
remnote_create_note |
Create a new note with title, content, parent, and tags |
remnote_search |
Search the knowledge base with query and filters |
remnote_read_note |
Read a note's content and children by ID |
remnote_update_note |
Update title, append content, add/remove tags |
remnote_append_journal |
Add an entry to today's daily document |
remnote_status |
Check connection status |
The bridge control panel is accessible via the right sidebar:
- Locate the MCP icon in RemNote's right sidebar toolbar
- Click the icon to open the control panel in the sidebar
- The panel displays:
- Connection Status - Current WebSocket connection state
- Session Statistics - Counts of created notes, updates, journal entries, and searches
- Action History - Last 10 MCP actions with timestamps
- Recent Logs - Real-time activity log
- The panel remains visible while you navigate RemNote (non-blocking)
- Click the icon again to close the panel
The sidebar panel provides persistent monitoring of MCP connection and activity while you work in RemNote.
Once everything is connected, you can ask your AI assistant things like:
- Create a note about the meeting we just had
- Search my notes in RemNote for information about AI coding
- Add a journal entry: Finished the MCP integration today!
- Find all my notes tagged with 'Ideas' and summarize them
- Update my 'Reading List' note with this new book
AI Assistant (Claude Code/Desktop) ↔ MCP Server (HTTP) ↔ WebSocket :3002 ↔ RemNote Plugin (this repo) ↔ RemNote SDK
Component roles:
- RemNote MCP Server (separate repository) - Exposes MCP tools to AI assistants and manages WebSocket server
- RemNote Automation Bridge (this repository) - RemNote plugin that connects to the server and executes operations via RemNote SDK
For RemNote domain concepts relevant to bridge behavior, see:
# Install dependencies
npm install
# Run in development mode (hot reload)
npm run dev
# Build for production
npm run build
# The plugin zip will be created as PluginZip.zipPlugin won't connect:
-
Verify plugin settings in RemNote:
- WebSocket URL:
ws://127.0.0.1:3002(default) - Check that MCP Server is running
- WebSocket URL:
-
Check plugin console (RemNote Developer Tools):
Cmd+Option+I (macOS) Ctrl+Shift+I (Windows/Linux) -
Restart RemNote after changing settings
"Invalid event setCustomCSS" errors:
- Cosmetic errors from development mode
- Don't affect functionality
- Won't appear in production builds
Notes not appearing:
- Check if a default parent ID is set (might be creating under a specific Rem)
- Verify the auto-tag setting isn't filtering your view
For server-related troubleshooting (installation, configuration, port conflicts, MCP tools not appearing in AI assistant), see the RemNote MCP Server documentation.
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
MIT License - see LICENSE for details.