Bear

Provides access to Bear notes for search, retrieval, and creation via MCP with local runtime configurations.
  • javascript

1

GitHub Stars

javascript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "philgetzen-bear-mcp": {
      "command": "node",
      "args": [
        "/path/to/bear-mcp/dist/index.js"
      ]
    }
  }
}

You can run a Bear MCP Server to let Claude Desktop read, search, and create Bear notes through the MCP (Model Context Protocol). It exposes your Bear data to Claude in a safe, structured way, enabling seamless note retrieval, tagging, and modification workflows.

How to use

You will run the local MCP server, then configure Claude Desktop to connect to it. With the server up, you can search notes, list tags, read note content, create new notes, and append text to existing notes. For best reliability, use simple search terms and retry once if a search times out.

How to install

Prerequisites you need before starting:"Node.js 18+" is required. Bear app should be installed on your system. Claude Desktop is needed to connect to the MCP server.

Install and build steps you should follow exactly as shown below to prepare the MCP server for use.

# 1) Clone the project
git clone https://github.com/philgetzen/bear-mcp.git
cd bear-mcp

# 2) Install dependencies
npm install

# 3) Build the project
npm run build

Configuration and usage notes

Two explicit runtime configurations are provided for running the MCP server locally. These configurations show how Claude Desktop will start the server using a local node process that runs the built index. You can choose the macOS path or the Windows path depending on your environment.

Configure Claude Desktop to load the Bear MCP server by adding one or both configurations to your claude_desktop_config.json. The entries below are ready to copy and adapt to your system paths.

{
  "mcpServers": {
    "bear": {
      "command": "node",
      "args": ["/path/to/bear-mcp/dist/index.js"]
    }
  }
}

Bear-specific setup for token access

Enable Bear x-callback-url access and generate an API token for search and tag operations. You will paste the token into Claude when prompted by the workflow.

Troubleshooting

If you encounter timeouts during searches, retry once with a simpler term. If browser windows briefly appear, they are part of Bear’s callback system and can be closed manually if they don’t disappear automatically. If Bear isn’t found, ensure Bear is installed and opened at least once.

For setup checks or diagnosing configuration problems, run a diagnostic command in Claude to verify Bear access and token validity.

Technical details

The server communicates via a local runtime using a standard process invocation. Typical response times are in the 1–5 second range for most operations, with longer times for searches and tag listing. This setup uses a local Node.js process that executes the built MCP server file.

Development

To run in development mode, you can start the server with the development script if provided by the project maintainers. Typical steps involve installing dependencies and building before starting the local server process.

Available tools

create_note

Creates a new Bear note and returns the note ID. This operation is reliable and messages the system with the title and content you provide.

add_text

Appends text to an existing Bear note identified by its ID. This operation is reliable for updating note bodies.

check_bear_setup

Runs a diagnostic check to verify Bear integration, including token validity and basic capability checks.

search_notes

Searches Bear notes and returns results with metadata. The reliability is high but may timeout in rare cases.

get_tags

Retrieves all tags from your Bear database. Useful for organizing and filtering notes.

get_note

Retrieves the full content of a note by its ID for reading and analysis.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational