Figure out structure of python code

This came up in a conversation with Victor from VITO:

* in our current Studio implementation, the python code contains 'hard-coded' boilerplate for the connection
* the OpenEO LLM Server returns the code with the generated connection

Figure out how to separate the boilerplate from the unique code to be used by the studio

Start by locating the current Studio implementation that emits Python and compare its hard-coded connection boilerplate with the code returned by the OpenEO LLM Server. Trace how the generated code is consumed by the Studio; done means the boilerplate and unique code have a documented separation suitable for the Studio.

Written by the indexing model from the issue text.