ZhongYao MCP Server

Project Overview

zhongyao-mcp-server is an AI tool server based on the Volcano Engine Doubao large model API and the MCP (Model Context Protocol) protocol. It focuses on the field of traditional Chinese medicine, providing a series of powerful multimodal content generation tools. Through this server, users can easily generate complete short videos containing detailed information, images, voice explanations, and video for specified Chinese herbs in MCP-supported clients (such as Trae, Cherry Studio) with a one-click operation.

The core of the project is to encapsulate complex, multi-step, multimodal generation processes (including data retrieval, copywriting, text-to-speech, video generation, and post-production) into simple and easy-to-use tools, greatly simplifying the creation process of TCM popular science content.

Core Functionality:

• Text-to-Image: Generate high-quality images based on text descriptions
• Text-to-Speech: Generate TTS voice broadcasts based on text
• Text-to-Video: Generate video content based on text descriptions
• Image-to-Video: Generate dynamic videos based on images and text descriptions
• Model Configuration: Support for multiple Doubao AI models

Supported AI Models:

• Language Model: deepseek-v3-250324
• Text-to-Image Model: doubao-seedream-3-0-t2i-250415
• Image-to-Video Model: doubao-seedance-1-0-lite-i2v-250428
• Text-to-Video Model: doubao-seedance-1-0-lite-t2v-250428
• Text-to-Speech: edgetts (free)

Deployment Guide

Environment Dependencies

• Python >= 3.11
• Volcano Engine Doubao API Key

This MCP dependency package is managed using uv, so the pyproject.toml file is:

json
dependencies = [
"mcp[cli]>=1.9.4", # Add requests dependency
"requests>=2.31.0",
"openai>=1.86.0",
"cos-python-sdk-v5==1.9.33",
"moviepy==1.0.3",
"httpx",
]

Configuration File Explanation

We use config.ini for the configuration file, with the following content:

ini

config.ini

[API]
api_key = YOUR_API_KEY_HERE
base_url = https://ark.cn-beijing.volces.com/api/v3

[Models]

Language Model

chat_model = deepseek-v3-250324

Text-to-Image Model

image_model = doubao-seedream-3-0-t2i-250415

Text-to-Video Model

video_model = doubao-seedance-1-0-lite-t2v-250428

[edgetts]
tts_api_key =zhouhuizhou
tts_base_url=https://edgettsapi.duckcloud.fun/v1

[common]
cos_region = ap-nanjing # Tencent Cloud OSS Storage Region
cos_secret_id = AKID003XXXXXXgO9qPl # Tencent Cloud OSS Storage SecretId
cos_secret_key = IZhavCXXXXXXXXXXX6i9NXUFqGTUOFvS # Tencent Cloud OSS Storage SecretKey
cos_bucket =tts-1258720957 # Tencent Cloud OSS Storage Bucket

font_path = /usr/share/fonts/truetype/wqy/wqy-zenhei.ttc

[Windows User Configuration]

imagemagick_binary = D:\develop\ImageMagick-7.1.1-Q16\magick.exe
font_path = SimHei

[Linux User Configuration]

imagemagick_binary = /home/ImageMagick/magick

font_path = /usr/share/fonts/truetype/wqy/wqy-zenhei.ttc

You need to modify the Volcano Engine api_key and Tencent Cloud OSS storage settings. Replace imagemagick_binary with the path where your program is installed.

For the container version (config.docker.ini), you only need to modify the Volcano Engine api_key and Tencent Cloud OSS storage settings; the imagemagick binary is pre-installed in the container image (no need to consider it).

Installation Methods

Method 1: Using pip

bash
pip install zhongyao-mcp-server

Method 2: Using uvx

bash
uvx zhongyao-mcp-server

Method 3: Using Docker (Recommended)

In the current directory, run the following command:

shell
docker build -t zhongyao_mcp_server .

After building the image, you can view the container image:

Start and run the container:

shell
docker run -d -p 8003:8003 --name zhongyao_mcp_server_instance zhongyao_mcp_server

Client Configuration

trae Configuration

Add the following configuration. Be sure to replace DOUBAO_API_KEY with your actual Volcano Engine Doubao API key.

Configuration using a configuration file in studio:

json
{
"mcpServers": {
"zhongyao-mcp-server": {
"name": "Chinese Medicine AI Video Generation Service",
"description": "Provides services for querying Chinese medicine information, generating copy, voice, and video, and final synthesis.",
"command": "uvx",
"args": [
"zhongyao-mcp-server"
],
"env": {
"ini": "F:\work\code\AIcode\zhongyao_mcp_server\config.ini"
}
}
}
}

Loading from environment variables in studio:

json
{
"mcpServers": {
"zhongyao-mcp-server": {
"name": "Chinese Medicine AI Video Generation Service",
"description": "Provides services for querying Chinese medicine information, generating copy, voice, and video, and final synthesis.",
"command": "uvx",
"args": [
"zhongyao-mcp-server"
],
"env": {
"DOUBAO_API_KEY": "Volcano Engine APIKEY",
"DOUBAO_BASE_URL": "https://ark.cn-beijing.volces.com/api/v3",
"DOUBAO_CHAT_MODEL": "deepseek-v3-250324",
"DOUBAO_IMAGE_MODEL": "doubao-seedream-3-0-t2i-250415",
"DOUBAO_VIDEO_MODEL": "doubao-seedance-1-0-lite-t2v-250428",
"TTS_API_KEY": "Custom TTS Password",
"TTS_BASE_URL": "https://edgettsapi.duckcloud.fun/v1",
"COS_REGION": "ap-nanjing",
"COS_SECRET_ID": "AKID0036B**********zSI1VjgO9qPl",
"COS_SECRET_KEY": "IZhavCL********88XUFqGTUOFvS",
"COS_BUCKET": "tts-1258720957"
}
}
}
}

sse configuration:

json
{
"mcpServers": {
"zhongyao-mcp-server": {
"url": "http://192.168.1.3:8003/sse"
}
}
}

Cherry Studio Configuration

1. Open Cherry Studio.
2. Go to Settings → MCP Servers → Add Server.
3. Configure the parameters:
   • Name: zhongyao-mcp-server
   • Description: Chinese Medicine AI Generation Service
   • Type: STDIO
   • Command: uvx
   • Arguments: zhongyao-mcp-server
   • Environment Variables: ini (the program depends on config.ini)
4. Click Save and Enable.

Detailed illustration:

sse configuration:

After configuring, you can check which tools are available:

stdio configuration:

Cross-Platform Considerations

To ensure smooth operation across different operating systems (such as Linux and Windows), please note the following:

1. Configuration File config.ini:
   • After packaging the project, the config.ini file will not be automatically installed in the system path. You need to manually copy config.docker.ini to the directory where you run the program and rename it to config.ini.
   • Modify the configuration items in config.ini according to your actual environment, especially the API keys and paths of external tools.
2. ImageMagick and FFmpeg:
   • The moviepy library depends on external tools ImageMagick and FFmpeg to handle videos and images. These tools need to be installed separately and are not distributed with the Python package.
   • Installation: Please download and install ImageMagick and FFmpeg from their official websites based on your operating system.
     • ImageMagick: https://imagemagick.org/* Configuration: After installation, make sure to correctly configure the imagemagick_binary path in the [common] section of the config.ini file. For example:
   • Windows: imagemagick_binary = D:\develop\ImageMagick-7.1.1-Q16-HDRI\magick.exe (please replace with your actual installation path)
   • Linux: imagemagick_binary = /usr/bin/magick (if added to PATH, the full path may not be necessary)

• If imagemagick_binary is not configured, MoviePy will attempt to find it in the system PATH. It is recommended to explicitly configure this to avoid potential issues.

API Key Acquisition

1. Visit Volcengine Console

2. Register and log in to your account

3. Activate the DouBao Large Model Service (you need to authorize each model separately)

   For Volcengine models, you need to authorize each one separately. Click on "Activate Management"

   

4. Create an API key in API Management

   API Management

   

Available Tools

2. get_chinese_herb_info

Retrieve detailed information about Chinese herbs.

• herb_name (string): Name of the Chinese herb.
• model (string, optional): The name of the model used to retrieve information, defaults to DEFAULT_CHAT_MODEL.

3. get_chinese_herb_image

Generate an image based on the name of a Chinese herb.

• herb_name (string): Name of the Chinese herb.
• size (string, optional): Image size, default is "1024x1024".
• model (string, optional): The name of the model used for generating images, defaults to DEFAULT_IMAGE_MODEL.
  Supported image sizes: 512x512, 768x768, 1024x1024, 1024x1792, 1792x1024

4. get_chinese_herb_video

Generate a video based on the name of a Chinese herb.

• herb_name (string): Name of the Chinese herb.
• duration (string, optional): Video duration (in seconds), default is "8".
• ratio (string, optional): Video aspect ratio, default is "16:9".
• model (string, optional): The name of the model used for generating videos, defaults to DEFAULT_VIDEO_MODEL.
  Supported video ratios: 16:9, 9:16, 1:1

5. generate_audio_from_text

Convert text to speech and upload it to COS.

• text (string): Text to be converted into speech.
• voice (string, optional): Voice model, default is "zh-CN-XiaoxiaoNeural".
• speed (float, optional): Speech playback speed, default is 1.0.

6. generate_herb_short_video

Create a complete short video about a specific Chinese herb, including information retrieval, script generation, voice synthesis, and video synthesis.

• herb_name (string): Name of the Chinese herb.

Usage Examples

1. Example of Retrieving Chinese Herb Information

bash
Use the get_chinese_herb_info tool to obtain detailed information about "Ginseng"

2. Example of Generating a Chinese Herb Image

bash
Use the get_chinese_herb_image tool to generate an image of "Goji Berry"

3. Example of Generating a Chinese Herb Video

bash
Use the get_chinese_herb_video tool to generate an 8-second video of "Angelica Sinensis"

4. Example of Converting Text to Speech

bash
Use the generate_audio_from_text tool to convert "Hello, this is a test voice." into speech

5. Example of Generating a Short Video About a Chinese Herb

bash
Use the generate_herb_short_video tool to create a short video about "Notoginseng"

Precautions

• Video generation tasks may take a long time to complete, please be patient
• Ensure that the API key has sufficient quota
• The URLs of generated content are time-limited, please save them promptly

Troubleshooting

Common Issues

1. API Key Error: Make sure the API key is correct and valid
2. Network Connection Issue: Check network connection and firewall settings
3. Model Unavailable: Confirm that the model name being used is correct

Debug Mode

Enable detailed log output:

bash
uvx zhongyao-mcp-server --verbose## Project Information

License: MIT License

Author: zhouqingyu666 Email: zhouqingyu666@aisino.com

Version: v0.0.5

• Initial version release
• Supports text-to-image, text-to-video, image-to-video, text-to-speech, and short video generation for traditional Chinese medicine
• Integrates ByteDance Volcano Engine API and edgetts

Contributions: Welcome to submit Issues and Pull Requests to improve this project