The problem here arises from a typo in the gammafit.pkg
file, which refers to a gammafit_lf.hlp which does not
exist. (The file which does exist is gammafit_lf.ado.)
I'll ask Kit Baum to fix the .pkg as soon as convenient.
This fix was applied last night (EST). The Web listings through IDEAS
and EconPapers refer to the wrong file today, but will sync tonight.
The "SSC" command in Stata works properly with this package (and is the
preferred method of accessing its contents). Sorry for the confusion.