Beemcp

🚀 BeeMCP Documentation

BeeMCP is a context manager based on the Model Context Protocol (MCP), designed to enhance the capabilities of AI assistants. It supports integration with various LLMs (Large Language Models) through the MCP protocol, enabling efficient management and retrieval of conversation history, user facts, to - do lists, and location data.

🚀 Quick Start

BeeMCP provides a command - line tool for users to quickly start using it:

Start the BeeMCP service

bee-mcp start

Access the BeeMCP management interface

After starting, you can access the management interface via the browser at http://localhost:3001.

✨ Features

BeeMCP offers a rich set of functional modules to help users better manage and retrieve data:

Conversation history management

• List all conversations: bee://conversations
• Get details of a specific conversation: bee://conversations/{id}

User fact management

• Record user facts: Use the record - user - fact tool to create new facts.
• Update existing facts: Use the update - user - fact tool to modify existing facts.
• Delete specified facts: Use the delete - user - fact tool to remove unwanted facts.

To - do list management

• List all to - do items: bee://todos
• Get incomplete to - do items: bee://todos/incomplete
• Get details of a specific to - do item: bee://todos/{id}

Location data management

• Record the current location: Use the record - location tool to record the geographical location automatically or manually.
• Query historical locations:
  • Locations in the past day: bee://locations/today
  • Locations in the past week: bee://locations/week
  • Locations in the past month: bee://locations/month

📦 Installation

BeeMCP offers multiple installation methods to meet the needs of different development environments:

Install using uv

npm install bee-mcp -g

Install using pip (for Python environments)

pip install beemcp

💻 Usage Examples

Basic Usage

Here are some example questions for using BeeMCP:

• "What were the conversation records from yesterday?"
• "Find the conversation content about the project with John."
• "Remember that I like black coffee. (-> record - user - fact)"
• "Actually, I like to add milk to my coffee. (-> update - user - fact)"
• "What are my to - do items?"
• "Show the list of incomplete tasks."
• "Add 'Buy groceries' to the reminder."
• "Mark the 'Send report' task as completed."
• "Where did I go last week?"
• "Which places have I visited today?"

Advanced Usage

Data Persistence

BeeMCP supports persisting data to local storage to ensure data is not lost after the service restarts.

Extended Plugin System

Users can develop custom plugins according to their needs to extend the functionality of BeeMCP. Plugins can be installed and deployed via the npm or pip package managers.

📚 Documentation

Debugging Tools

To facilitate developers' debugging, BeeMCP provides the MCP inspector tool:

Run using uvx

npx @modelcontextprotocol/inspector uvx beemcp

Install using pip

npx @model