Quick start#
Learn how to connect PyLumerical-MCP to your preferred agentic AI client. Before you begin, install PyLumerical-MCP by following the instructions in Installation.
Connect to your agentic client#
You can use PyLumerical-MCP with transport over both Streamable HTTP and STDIO.
Since PyLumerical-MCP uses the open source MCP standard, which enables AI applications to seamlessly integrate with external systems, you can use it with a wide range of agent-based frameworks and agent harnesses that support MCP.
The following sections show you the important settings for PyLumerical-MCP. For instructions on how to connect, see your preferred client’s documentation.
Configure Streamable HTTP transport#
To configure Streamable HTTP transport, you can set it through the .env file.
Copy the
.env.exampletemplate file and rename it to.env.Set the following variables:
FASTMCP_TRANSPORT: Set to"streamable-http".FASTMCP_HOST: Set to the hostname for the server. For local use, set to"127.0.0.1".FASTMCP_PORT: Set to the port for the server, such as8081.
Start the MCP server by running
ansys-lumerical-mcpfrom the command line in the environment where the MCP server is installed.Connect to the MCP server from your client.
Configure STDIO transport#
The configuration for STDIO transport depends on your specific client. However, these general guidelines can help:
Set the target executable to the
ansys-lumerical-mcpin the environment where it is installed.Set the transport type to
stdio. The specific key can vary by client.Leave the argument field empty.
View additional resources#
View example prompts for using PyLumerical-MCP to drive Lumerical tools.
Learn how to use PyLumerical-MCP effectively.
Access references for every tool the server exposes.