When creating archives for IGOs datasets, we have realized we were to add a code to the datasets (notably to identify the zip file). There might be some thinking to be done around this (How do we determine the codes to be used? Etc.)
I think in this case the code would only be for your use and to facilitate your work. You should define a syntax which guarantees that there is no risk of having the same code as an OECD dataset.
Otherwise, you choose the code you prefer to recognize the dataset.
I do not think this needs a development on our side, just a decision you take and you can let us know