Thanks for this Evan! Traditionally we don't add the lut.nc files - we could (it's small), but it's probably a benefit for the model to reconstruct it every time on the fly. Is that rationale for including to abstract out the requirement for the 6c emulator while it's still developmental?

Everything else looks good to me!

@pgbrodrick I included the .nc file because I actually wasn't sure if the actions had access to sRTMnet (or which version it had access to). Is it correct that the .sh, .inp, and 6s outputs are all 6s-related files and that without the .nc file, the sRTMnet predictions would have to still run?

If that is the case and I'm not completely wrong, I added the .nc assuming that we weren't planning on linking the development 6c emulator into the actions.

I can add/remove the .nc files as needed so any guidance w.r.t. consistency is appreciated.