I think this all looks okay. The script will align the anat to the epi dataset by default including matching the epi dataset's obliquity. If the input time series has slice timing information, the script will use that info. 3dTshift can be used to supply slice timing info to the dataset if it's missing for some reason, but that should be a relatively rare occurrence (but it looks like it could be a problem for you). Likewise, the script will check for obliquity, and if it exists, it will use that. You can always call the align_epi_anat.py script from another script that fixes the input files. This previous thread discusses a couple ways to supply slice timing information for a dataset:
[
afni.nimh.nih.gov]