mcp-server

简体中文

Huawei MCP Server is a Model Context Protocol server built on Huawei Cloud services, providing secure and controlled cloud access capabilities for large AI models. Through standardized MCP specifications, it enables AI assistants to operate Huawei Cloud resources within conversational workflows, supporting core services including ECS, OBS, GaussDB, and other widely-used cloud products.

Mcp-server catalog

• Practical documents
• Demo
• Running Guide
• Tools list
• Contributing

Practical documents

• Using DAS MCP to implement database conversational operation and maintenance

Demo

Demo1

Cline uses Huawei Cloud EIP MCP Server to create and delete EIP instances.

Running Guide

1. Dependency Installation

Install the Python environment in advance. Since Python 3.4 and 2.7.9, pip has been installed with Python as a standard component.

• Install uv (Recommended)
• Install python version 3.10 or above

2. Environment variable settings

Prepare AK and SK and set them to environment variables

• ak environment variable name: HUAWEI_ACCESS_KEY
• sk environment variable name: HUAWEI_SECRET_KEY

3. Running method

Take running the mcp-server-ecs service as an example

Run with 'uv' (recommended)

# Enter the root path of the project
cd /path/to/you/mcp-server


# Start the service
uv run mcp-server-ecs

Execute uv run mcp-server-ecs -h to view the usage instructions, the configuration values of the config.yaml file in the sub-project can be overwritten by using the optional parameters in the command line

usage: mcp-server-ecs [-h] [-p PORT] [-t {http,sse,stdio}]

MCP Server

options:
  -h, --help            show this help message and exit
  -p, --port PORT       Port number
  -t, --transport {http,sse,stdio}
                        Transport of MCP Server

Run with python

# Enter the root path of the project
cd /path/to/you/mcp-server


# Install dependencies
pip install -e .


# Enter the specified service (mcp-server-ecs) path
cd huaweicloud_services_server/mcp_server_ecs/src/mcp_server_ecs


# Start the service
python run.py

MCP Marketplace Integration

• Cline
• Configure the mcp service to use http in cline. The json format is as follows

{
  "mcpServers": {
    "mcp_server_ecs": {
      "url": "http://localhost:8888/mcp",
      "disabled": false,
      "type": "streamableHttp",
      "autoApprove": []
    }
  }
}

Tools

Contributing

We welcome contributions from the open-source community! If you'd like to contribute to this project, please refer to the contributing guide.