
Bear MCP Server
An MCP server that integrates Bear Note Taking App with Claude Desktop, allowing Claude to read, create, search notes and manage tags directly from Bear.
README
Bear MCP Server
An MCP (Model Context Protocol) server for integrating Bear Note Taking App with Claude Desktop. This server allows Claude to read, create, and search your Bear notes directly.
✨ Features (v4.0.2)
Full data retrieval capabilities:
- 🔍 Search notes and get complete results with metadata
- 🏷️ Retrieve all tags from your Bear database
- 📖 Read note content for analysis and summarization
- ✏️ Create notes and get their IDs back
- 📝 Add text to existing notes
- ✅ Test setup with comprehensive status checking
⚠️ Current Limitations & Usage Guidelines
Browser Window Behavior
Expected: Brief browser windows may appear during search/tags operations due to Bear's callback system.
- Windows are automatically minimized and moved off-screen
- Auto-close within 1-2 seconds in most cases
- Safe to manually close if they persist
- Focus is returned to your original window automatically
Search Reliability
Success Rate: ~80-90% of searches work reliably
- Occasional timeouts may occur (20-second limit)
- Simple terms work better than complex queries
- Retry once if a search times out
- Single words tend to be more reliable than phrases
Best Practices
✅ Recommended Usage:
Search my Bear notes for "project"
Get all my Bear tags
Create a note titled "Meeting Notes" with today's agenda
Add "Action item: Follow up" to my "Weekly Review" note
⚠️ If Issues Occur:
- Timeouts: Wait a moment and retry the same search
- Browser windows: Safe to close manually if they don't auto-close
- No results: Try simpler/broader search terms
- Setup issues: Use
check_bear_setup
to diagnose
Available Tools
Tool | Functionality | Reliability | Notes |
---|---|---|---|
create_note |
Creates new notes | ✅ Excellent | Always works, returns ID |
add_text |
Adds text to notes | ✅ Excellent | Reliable text addition |
check_bear_setup |
Tests configuration | ✅ Excellent | Diagnostic tool |
search_notes |
Searches notes | ⚠️ Good | ~80-90% success, may timeout |
get_tags |
Lists all tags | ⚠️ Good | Usually works, brief popup |
get_note |
Retrieves note content | ⚠️ Good | Works well with valid IDs |
Installation
- Clone the repository:
git clone https://github.com/philgetzen/bear-mcp.git
cd bear-mcp
- Install dependencies:
npm install
- Build the project:
npm run build
Configuration in Claude Desktop
Add the server to your Claude Desktop configuration file:
macOS
Edit ~/Library/Application Support/Claude/claude_desktop_config.json
:
{
"mcpServers": {
"bear": {
"command": "node",
"args": ["/path/to/bear-mcp/dist/index.js"]
}
}
}
Windows
Edit %APPDATA%\Claude\claude_desktop_config.json
:
{
"mcpServers": {
"bear": {
"command": "node",
"args": ["C:\\path\\to\\bear-mcp\\dist\\index.js"]
}
}
}
Bear Configuration
Required Settings:
-
Enable x-callback-url:
- Open Bear → Settings (⌘,)
- Go to "Advanced" tab
- Enable "Allow x-callback-url"
-
Generate API Token (for search and tags):
- Open Bear → Help → Advanced → API Token
- Click "Copy Token"
- In Claude, use:
set_bear_token
with your token
Usage Examples
Setup and Testing
Check my Bear setup and show me what's available
Reliable Operations (Always Work)
Create a new Bear note titled "Project Ideas" with content about machine learning
Add "## Next Steps\n- Review documentation\n- Schedule follow-up" to my "Weekly Review" note
Search Operations (Usually Work)
Search my Bear notes for "machine learning" and show me what you find
Get all my Bear tags and help me organize them
Content Analysis (When Search Works)
Search for "meeting notes" then help me identify common action items across all results
Find notes tagged with "work" and summarize the main topics
Troubleshooting
"Search failed: Callback timeout"
- Normal: Happens ~10-20% of the time
- Solution: Wait 5-10 seconds and retry the same search
- Tip: Try simpler search terms (single words work better)
"No token configured"
Get your token: Bear → Help → Advanced → API Token → Copy Token, then use set_bear_token
"Bear not found"
Make sure Bear is installed and has been opened at least once.
Browser windows appearing
- Expected behavior due to Bear's callback system
- Windows auto-minimize and move off-screen
- Safe to manually close if they persist
- Focus returns to your original window automatically
No search results
- Check search term: Ensure it exists in your notes
- Verify token: Use
check_bear_setup
to test - Try broader terms: Single words often work better than phrases
- Check Bear directly: Verify the content exists in Bear app
Technical Details
HTTP Callback System
The server uses HTTP callbacks (port 51234) to receive data from Bear:
- Bear sends search results and tag data via URL callbacks
- Browser windows appear briefly due to this callback mechanism
- 20-second timeout for Bear responses
- Auto-retry mechanisms for common failures
Performance Characteristics
- Fast operations: Note creation, text addition (~1-2 seconds)
- Medium operations: Single note retrieval (~3-5 seconds)
- Slower operations: Search, tags list (~5-15 seconds)
- Timeout threshold: 20 seconds maximum wait
Development
To run in development mode:
npm run dev
To test the server:
npm run build
node dist/index.js
Version History
- v4.0.2: Enhanced browser window handling, 20s timeout
- v4.0.1: Improved callback reliability, better error messages
- v4.0.0: Full functionality restored with callback system
- v3.1.0: Documentation-only version (limited functionality)
Requirements
- Node.js 18 or higher
- Bear app installed on your system
- Claude Desktop application
- macOS (Bear is macOS-only)
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.