I suggest providing a function to create a tar archive with all files that are needed for the custom viewer, which can be used by Tools that need it.
It could be called within Tool.run() specifying the files that need to be archived, e.g.:
custom_view_archive(file_list, archive_name)
Alternatively it could be called automatically by App when output_files with "data_type": "tool_statistics" are present; note that this requires specifying a priori a list of files to be archived (e.g. in Tool.json).
I suggest providing a function to create a tar archive with all files that are needed for the custom viewer, which can be used by Tools that need it.
It could be called within Tool.run() specifying the files that need to be archived, e.g.:
custom_view_archive(file_list, archive_name)Alternatively it could be called automatically by App when output_files with "data_type": "tool_statistics" are present; note that this requires specifying a priori a list of files to be archived (e.g. in Tool.json).