Hi Alesventus,

The approach suggested by syzmon_dybczak is definitely the same direction I would go in. To supplement their answer, have a look at this Databricks Labs repository that you may find useful as inspiration (DLT-Meta): https://github.com/databrickslabs/dlt-meta.

I have referenced this multiple times in the past and whilst this specific repository specifically looking at Lakeflow Declarative Pipelines, the patterns presented are agnostic to the tool. I think you may find some reusable patterns in the structure of the suggested control table structure and how you may find that they interact with your specific use case.