Openeuler MCP Servers

🚀 openEuler MCP Servers Repository

The openEuler MCP Servers repository provides a common protocol for large - model contexts, enabling calls to various applications and enhancing the capabilities of large models. It focuses on the operating system domain, integrating with Devstation and EulerCopilot to improve the openEuler interaction experience.

✨ Features

MCP, short for Model Context Protocol, aims to offer a general - purpose large - model context protocol. It allows for the invocation of various applications and the expansion of large - model capabilities. The openEuler mcp - servers repository is dedicated to storing various MCP Servers, with a focus on the operating system domain. It collaborates with Devstation and EulerCopilot to enhance the openEuler interaction experience.

🔧 Technical Details

Software Architecture

The repository directory includes a doc directory for storing various document introductions. Each independent MCP Server is located under the servers folder:

mcp-servers/
├── servers
|      |──oeDeploys
|           |── readme.md
|           |── mcp_config.json
|           └── src
|      |──oeGitExt 
|      └──xxxxx
└── doc

Taking oeDeploy as an example, each MCP Server top - level should contain the following:

1. A description of the functions implemented by the MCP Server and how to utilize its capabilities.
2. A JSON file for MCP configuration.
3. Source code.

🚀 Quick Start

1. Environment Setup

• Install DevStation and start VSCodium.
• Install Roo Code from the plugin market (EulerCopilot can be used as a replacement in the future).

2. Configure the Large - Model Service

Configure large - model parameters in the Roo Code settings:

{
  "apiProvider": "OpenAI Compatible",
  "openaiBaseUrl": "https://api.siliconflow.cn",
  "key": "your_api_key",
  "model": "Pro/deepseek-ai/DeepSeek-V3"
}

3. Create an MCP Service

Write an MCP Server script (e.g., oegitext_mcp.py):

from mcp.fastapi import FastAPI, setup_mcp_router

app = FastAPI()
setup_mcp_router(app)

if __name__ == "__main__":
    app.run()

4. Configure the MCP Client

Edit the global MCP configuration file and add a new MCP Server:

{
  "mcpServers": {
    "oegitext_mcp": {
      "command": "/home/xxx/.local/bin/uv",
      "args": [
        "--directory",
        "/home/xxx/oegitext_mcp",
        "run",
        "oegitext_mcp.py"
      ],
      "disabled": false,
      "autoApprove": [],
      "alwaysAllow": []
    },
  }
}

5. Use the MCP Service

Call the configured MCP Server through Roo Code or EulerCopilot to achieve large - model - enhanced code completion and interaction experience.

💻 Usage Examples

Basic Usage

1. Use common MCP clients (such as EulerCopilot, Roo Code, VSCodium, etc.) for calls.
2. Configure MCP server information globally or at the project level.

For example, edit the global MCP configuration in the Roo Code plugin:

{
  "mcpServers": {
    "oegitext_mcp": {
      "command": "/home/xxx/.local/bin/uv",
      "args": [
        "--directory",
        "/home/xxx/oegitext_mcp",
        "run",
        "oegitext_mcp.py"
      ],
      "disabled": false,
      "autoApprove": [],
      "alwaysAllow": []
    },
  }
}