These should be reconfigured to behave similarly to how cube and deck images are. Rather than specifying the url to the image itself, which scryfall's documentation says to avoid as the locations may change over time, the image field on the back end should store the card's scryfall_id (or possibly the card's set code and collector number; pending a response from the scryfall team regarding collection data language requests). Then when the data is fetched from CLM's server, required requests are sent to scryfall's servers to then cache on the CLM front end.
These should be reconfigured to behave similarly to how cube and deck images are. Rather than specifying the url to the image itself, which scryfall's documentation says to avoid as the locations may change over time, the image field on the back end should store the card's scryfall_id (or possibly the card's set code and collector number; pending a response from the scryfall team regarding collection data language requests). Then when the data is fetched from CLM's server, required requests are sent to scryfall's servers to then cache on the CLM front end.