Xiaozhi MCP Aggregate

Provides an interface to extend AI capabilities with tools for Dingtalk, email, system status, and web search via MCP.

python

49

GitHub Stars

python

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": {
    "avxxoo-xiaozhi-mcp": {
      "command": "python",
      "args": [
        "mcp_pipe.py",
        "aggregate.py",
        "--env-file",
        ".env.xiaozhi1"
      ],
      "env": {
        "YOUR_ENV_VAR": "<YOUR_ENV_VAR>"
      }
    }
  }
}

You set up and run XiaoZhi MCP to extend AI capabilities with tools you can invoke from models, such as sending emails, performing web searches, or querying server status. This server-based approach lets your AI interact with external systems in real time and under your control.

How to use

To use this MCP server, connect with a compatible MCP client and enable tool invocation from your AI workflows. You will access tools by their defined names, each exposing a clear interface (inputs and outputs) that the model can call. Tools support remote control, calculations, email operations, knowledge search, and more, enabling your AI to interact with external systems through standardized endpoints.

How to install

Prerequisites you need before installation include Python 3.10 or newer and a Python virtual environment to keep dependencies isolated.

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Set up your configuration by creating a local environment file. Copy the example configuration to your own file, for example .env.xiaozhi1, and customize variables as needed.

# Copy the example to your own file and edit as needed
# Implementations typically include API keys, endpoints, and tool-specific settings

Run the aggregate MCP access point using your environment file. You can run different XiaoZhi MCP access points by specifying different configuration files.

python mcp_pipe.py aggregate.py --env-file .env.xiaozhi1

Notes on configuration and tool creation

If you want to create your own MCP tools, follow these practical steps to extend capabilities and register new tools.

Creating Your Own MCP Tools

Here's a simple guide to add a custom MCP tool and register it with the aggregate server.

Create a tool following the established naming convention, using a function_channel style to distinguish channels (for example, email_google for a Google Mail tool).

Register your tool in the aggregate configuration so the MCP server exposes it to models.

Configure any needed environment variables in a dedicated environment file (if applicable) and document them in your tooling references.

Available tools

conversation_dingtalk

Send conversation content to a DingTalk group bot as part of your AI workflow.

email_qq

Send emails using a QQ mail account from the MCP tools.

system

Retrieve server status information including CPU, memory, and disk usage.

web_webpilot

Perform web searches through the WebPilot integration for online information retrieval.