Taking a brief look at the paper, I’m not aware of those cm_scripts being part of the standard Rosetta distribution.
I might suggest contacting the authors to see if there’s a version of the scripts which are publically availible. Specifically, you may want to contact , who is now a professor at UCSC (http://niksgourakis.chemistry.ucsc.edu/). He is currently the maintainer of CSRosetta (https://csrosetta.chemistry.ucsc.edu/), a collection of scripts which is useful for setting up and running Rosetta with NMR data (and not just chemical shifts). I’m not familiar enough with CSRosetta to know off-hand, but there may be equivalents for the scripts you need in the current releases of CSRosetta.
Another thing I’ll mention is that the comparative modeling protocol outlined in that paper has been largely supplanted by the RosettaCM protocol (https://www.ncbi.nlm.nih.gov/pubmed/24035711). It should be possible to incorporate similar sorts of NMR data with the RosettaCM protocol, though the exact methods on how to do so have likely changed.
.
