I believe uc2fasta is only marginally useful: in effect, it did this: "retrieve the sequences with these labels from that FASTA file", and any bioinformatics installation should have a utility for that. Output files generated by -centroids and -matched provide FASTA output for the most common use cases.

If you have filtered a .uc using your own script to select a subset of records, then you can get just the query labels by using the cut command to extract column 9, like this:

  cut -f9 myresults.uc > labels.txt

The uc2fastax commands was intended for making ustar alignments, which is no longer needed because you can use -msaout to generate alignments directly. So uc2fastax is obsolete.

If you really need one of these these commands for some reason, then you can download a v5 binary by going to the usual download page. Notice the pull-down selection for older versions: