Xiaozhi Autoglm MCP

🚀 xiaozhi-AutoGLM-mcp

This is a Xiaozhi MCP service built on Open-AutoGLM. As an MCP service, it connects to Xiaozhi AI and xiaozhi-esp32-server and is invoked by Xiaozhi.

🚀 Quick Start

✨ Features

This service provides seamless integration with Xiaozhi AI and related servers, leveraging the power of Open-AutoGLM for efficient operations.

📦 Installation

Environment Preparation

1. ADB (Android Debug Bridge)

1. Download the official ADB installation package and extract it to a custom path.
2. Configure environment variables:

• MacOS: In Terminal or any command-line tool, run the following command:

# Assume the extracted directory is ~/Downloads/platform-tools. Adjust the command if necessary.
export PATH=${PATH}:~/Downloads/platform-tools

• Windows: Refer to the third-party tutorial for configuration.

2. Android 7.0+ Device or Emulator with Developer Mode and USB Debugging Enabled

1. Enable Developer Mode: Usually, you can find Settings - About Phone - Version Number and click it quickly about 10 times until a pop-up window shows "Developer mode has been enabled". The process may vary slightly for different phones. If you can't find it, search for a tutorial online.
2. Enable USB Debugging: After enabling developer mode, go to Settings - Developer Options - USB Debugging and check the box to enable it.
3. Some models may require a device restart after setting developer options to take effect. You can test it by connecting the phone to the computer via a USB data cable and running adb devices to check if device information is displayed. If not, the connection fails.

Please carefully check the relevant permissions:

3. Install ADB Keyboard (for text input)

Download the installation package and install it on the corresponding Android device. Note that after installation, you need to go to Settings - Input Method or Settings - Keyboard List to enable ADB Keyboard for it to work.

Deployment Preparation

1. Install Dependencies

conda remove -n autoglm --all -y
conda create -n autoglm python=3.10 -y
conda activate autoglm

pip install -r requirements.txt 

2. Configure ADB

Ensure that the USB data cable has data transfer capabilities, not just charging capabilities. Make sure ADB is installed and connect the device using a USB data cable:

# Check connected devices
adb devices

# The output should show your device, for example:
# List of devices attached
# emulator-5554   device

3. Configure Visual Model and Xiaozhi MCP Access Point

1. Copy the config.yaml file to .config.yaml.
2. Log in to Xiaozhi AI or your privately deployed Smart Control Console to obtain the MCP access point address of the agent.
3. Edit the .config.yaml file and replace the obtained MCP access point address in the mcp_endpoint field.
4. If you choose ChatGLMVLLM as the visual large language model in your .config.yaml, go to the Zhipu AI platform to obtain your API key. Make sure your account balance is sufficient; otherwise, the invocation will fail.
5. Edit the .config.yaml file and replace the obtained API key in the api_key field under ChatGLMVLLM in VLLM.

4. Run the Service

• Linux/macOS:

./start_mcp_service.sh

• Windows:

python mcp_pipe.py mobile_agent_server.py