Evaluating Results of Result Smoothing - Constant in Network Elements Using AI Agents / MCP Server

Good day,

as may have already become apparent from some of my other posts, I currently have problems accessing the results with the result smoothing setting “constant in finite elements” via the API and am therefore currently trying to find a kind of workaround.

One idea that occurred to me was to use an AI agent that runs on the MCP server; I read about this in the following community post: AI integration in parametric design

I would like to simply try it out myself; however, on the one hand, I have the problem that I cannot place the config.ini file in the folder path due to a security blockade of the API, and on the other hand, I fear that the AI agent ultimately only generates processes based on the Python API commands and that there is currently simply no way to access result values of the result smoothing “constant in finite elements”…

Since I do not want to waste more hours tomorrow training the AI agent and holding up other colleagues from IT only to realize that the agent cannot evaluate constant mesh smoothing, I thought I would simply ask in advance whether the AI agent actually just generates Python API codes or can also access more results such as those from other result smoothings / raw data / integration points, etc.

Best regards,

Nick Böttcher

Admin rights are required to edit files in the Program folder. Try starting Notepad with admin rights; then creating the config.ini should work.

It is as you suspect, the MCP server ultimately generates "only" Python code. What the API cannot do, the MCP server cannot do either.

You can think of the MCP server as an additional layer over the API that makes the API accessible to the language model.

Best regards
Frank

Thank you very much for the quick response!

1 Like