AVEVA MCP Server (Compiled Distribution)

A Model Context Protocol (MCP) server that exposes AVEVA process simulation tools to AI applications like Claude Desktop, Cursor, and other MCP-compatible clients.

⚠️ Requirements

Requirement	Version
Python	3.13.x (must match exactly)
OS	Windows 10/11 (64-bit)
AVEVA SDK	Installed separately

Important: The compiled .pyd files are built for Python 3.13. Using a different Python version will cause import errors.

🚀 Quick Start

1. Install Dependencies

pip install -r requirements.txt

Note: AVEVA dependencies must be installed separately from the AVEVA SDK.

2. Run the MCP Server

# Default: stdio transport (for Claude Desktop/Cursor)
python start_aveva_mcp_server.py

# SSE transport (for web-based clients)
python start_aveva_mcp_server.py sse --port 8000

# Streamable HTTP transport (for modern clients)
python start_aveva_mcp_server.py http --port 8000

If you run successfully, you can stop the server and move on to next step.

3. Configure Claude Desktop (MCP client)

You can find a detailed tutorial on how to locate and edit your Claude Desktop configuration file in the official documentation: Connect to local MCP servers For other MCP clients, please refer to their respective documentation on configuring local MCP servers. Although the interfaces may differ, the core logic remains the same: you need to specify:

• Specify the Python executable that will run the server
• Specify the path to the script that launches your MCP server

Add the following to your Claude Desktop configuration file. Update the paths to match your environment:

{
  "mcpServers": {
    "aveva": {
      "command": "C:/Users/xxxx/AppData/Local/miniconda3/envs/your_py313_env/python.exe",
      "args": [
        "C:/path/to/distribute/start_aveva_mcp_server.py"
      ]
    }
  }
}

Quit the claude desktop completely from the background, and restart it, the MCP tools will be automatically loaded when claude desktop starts. Try to ask something and explicitly mentioning using AVEVA tools.

You can see the claude agent used the tools "Aps connect" and "Sim create" for the task.

📁 Package Contents

File	Description
start_aveva_mcp_server.py	Main launcher script
aveva_mcp_server.py	MCP server with tool definitions
requirements.txt	Python dependencies
tools/*.pyd	Compiled logic modules
tools/__init__.py	Package initializer

🔧 Troubleshooting

"DLL load failed" or "ImportError"

• Ensure you are using Python 3.13.x (64-bit)
• Run python --version to verify

"ModuleNotFoundError: No module named 'fastmcp'"

• Run pip install -r requirements.txt

AVEVA connection errors

• Verify AVEVA Process Simulation is installed
• Ensure AVEVA SDK Python components are accessible
• Test with: python -c "import simcentralconnect; print('AVEVA OK')"

📚 Learn More

• Model Context Protocol
• FastMCP Documentation
• MCP Python SDK
• AVEVA Process Simulation Documentation

---

Citation

Liang, Jingkang, Niklas Groll, and Gürkan Sin. "Large Language Model Agent for User-friendly Chemical Process Simulations." arXiv preprint arXiv:2601.11650 (2026).