IDE and client configuration#

PyAEDT-MCP works with any MCP-compatible client. This page covers the most common ones: Claude Code, Visual Studio Code, and Claude Desktop.

Claude Code#

Claude Code is Anthropic’s AI-powered code editor with built-in MCP support.

Configure globally#

Configure PyAEDT-MCP for all your Claude Code projects:

claude mcp add --transport stdio --scope user pyaedt-mcp -- \
  uvx --from git+https://github.com/ansys/pyaedt-mcp ansys-aedt-mcp

Advantages:

  • Available in all Claude Code projects without per-project configuration.

See Claude Code MCP installation for details.

Visual Studio Code#

Visual Studio Code integrates MCP servers through the Copilot extension.

Set up for local development#

Use this configuration when working from a local clone of the repository:

{
  "servers": {
    "pyaedt-mcp": {
      "type": "stdio",
      "command": ".venv/Scripts/python",
      "args": ["-m", "ansys.aedt.mcp"],
      "env": {
        "FASTMCP_LOG_LEVEL": "DEBUG"
      }
    }
  }
}

Note

On Linux or macOS, use bin/python instead of Scripts/python.

Use uv as an alternative#

If you prefer, you can use uv as your Python package and project manager:

{
  "servers": {
    "pyaedt-mcp": {
      "type": "stdio",
      "command": "uv",
      "args": ["run", "python", "-m", "ansys.aedt.mcp"]
    }
  }
}

Configure HTTP transport#

If you start PyAEDT-MCP with --transport http, use this client configuration:

{
  "servers": {
    "pyaedt-mcp": {
      "type": "http",
      "url": "http://127.0.0.1:8080"
    }
  }
}

Start PyAEDT-MCP before you connect:

ansys-aedt-mcp --transport http --http-host 127.0.0.1 --http-port 8080

Use Docker endpoint#

If you start PyAEDT-MCP with Docker Compose, use the default HTTP endpoint:

{
  "servers": {
    "pyaedt-mcp": {
      "type": "http",
      "url": "http://localhost:8080"
    }
  }
}

For more information, see Docker deployment.

Enable MCP in Visual Studio Code#

  1. Open VS Code settings (Ctrl+, or Cmd+,).

  2. Search for MCP.

  3. Enable the settings that allow Copilot to use MCP servers.

    For more information, see Add and manage MCP servers in VS Code in the Visual Studio Code documentation.

  4. Restart Visual Studio Code.

Claude Desktop#

Claude Desktop is Anthropic’s macOS desktop app with full MCP support. Edit the ~/Library/Application Support/Claude/claude_desktop_config.json file on macOS or the equivalent path on Windows:

{
  "mcpServers": {
    "pyaedt-mcp": {
      "command": "uvx",
      "args": [
        "--from", "git+https://github.com/ansys/pyaedt-mcp.git",
        "ansys-aedt-mcp"
      ],
     "description": "MCP server for Ansys AEDT through PyAEDT",
     "version": "0.0.1",
     "language": "python"
    }
  }
}

Claude Code versus Visual Studio Code#

Feature

Claude Code

Visual Studio Code

Configuration method

CLI command (claude mcp add)

JSON file (.vscode/mcp.json)

Setup level

Project or global

Project-level only

Transport support

STDIO (default)

STDIO or HTTP

Team sharing

With project config files

With .vscode/mcp.json in the repository

Learning curve

Low (CLI-based)

Medium (JSON configuration)

Advanced configuration#

Auto-connect to AEDT on startup#

Pass --connect to have the server connect to AEDT during initialization. Use --machine and --port to target a specific gRPC endpoint:

Visual Studio Code

Edit the .vscode/mcp.json file:

{
  "servers": {
    "pyaedt-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "--index-strategy", "unsafe-best-match",
        "--from", "git+https://github.com/ansys/pyaedt-mcp.git",
        "ansys-aedt-mcp",
        "--connect",
        "--machine", "192.168.1.100",
        "--port", "50051"
      ]
    }
  }
}

Claude Code:

claude mcp add --transport stdio pyaedt-mcp -- \
  uvx --from git+https://github.com/ansys/pyaedt-mcp ansys-aedt-mcp \
  --connect --machine 192.168.1.100 --port 50051

Warning

When --connect is used, the server locks the connection. The launch_aedt, connect_to_aedt, and disconnect_from_aedt tools are disabled for the lifetime of the server process.

Enable optional context tools#

The --include-context flag registers get_guidelines_for, which provides inline AEDT and PyAEDT workflow guidance to the AI assistant.

{
  "servers": {
    "pyaedt-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "--index-strategy", "unsafe-best-match",
        "--from", "git+https://github.com/ansys/pyaedt-mcp.git",
        "ansys-aedt-mcp",
        "--include-context"
      ]
    }
  }
}

Enable dynamic tool discovery#

Use --dynamic-tool-discovery to hide AEDT-only tools until a session is established. This keeps the AI assistant’s context small before connection.

ansys-aedt-mcp --dynamic-tool-discovery

Enable debug logging#

Set the FASTMCP_LOG_LEVEL environment variable to DEBUG:

Visual Studio Code

Edit the .vscode/mcp.json file:

{
  "servers": {
    "pyaedt-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "--index-strategy", "unsafe-best-match",
        "--from", "git+https://github.com/ansys/pyaedt-mcp.git",
        "ansys-aedt-mcp"
      ],
      "env": {
        "FASTMCP_LOG_LEVEL": "DEBUG"
      }
    }
  }
}

Command line:

FASTMCP_LOG_LEVEL=DEBUG ansys-aedt-mcp

Connect with HTTP (Docker or remote)#

For containerized or remote deployments, see Docker deployment.

Next steps#