The datetime.now() function in find_missing_datasets.py generates a new timestamp every time the script runs, even if the content of the missing datasets report hasn't changed. This leads to unnecessary diffs in version control for the missing-datasets.md file. Consider either using a fixed placeholder or only updating the timestamp if the actual list of missing datasets changes.
Relevant code:
https://github.com/LDFLK/datasets/blob/main/find_missing_datasets.py\#L67-L72