This is an issue with JEFF-3.3 not following the ENDF-6 format, rather than an issue with ENDFtk, but it's worth keeping track of these.
Consider the decay file JEFF33RDD0506_0.ASC, as downloaded from https://www.oecd-nea.org/dbdata/jeff/jeff33/index.html -> Radioactive Decay Data in 3852 files. In the File 1 MT=451 dataset, the data is written as:
[...]
diff. nuclear -6.1026 keV -0.0587 % 506 1451 45
506 1451 46
0.00000+00 0.00000+00 1 451 53 0 506 1451 47
0.00000+00 0.00000+00 8 457 35 0 506 1451 48
506 1 099999
506 0 0 0
According to the ENDF-6 format, the first two 0.00000+00 values should not be there. ENDFtk attempts to parse them as an integer and fails.