As mentioned, if you have a .tbi file, then the prefix of that file with .gz probably means it was made with bgzip.

You can use htslib to check if the file is block-compressed, or use hexdump to check the first two bytes (for the file type), the fourth byte (to see if there is an extra header set) and the 13th and 14th bytes (to check for the extra header):

$ hexdump -s 0 -n 2 -e '8/1 "%02x""\n"' some_file.gz
1f8b
$ hexdump -s 3 -n 1 -e '8/1 "%d""\n"' some_file.gz | awk 'and($0,0x04){ print "extra header"; }'
extra header
$ hexdump -s 12 -n 2 -e '8/1 "%c""\n"' some_file.gz
BC

If the first result equals 1f8b, the second result returns extra header, and the third equals BC, then some_file.gz was probably made with bgzip.

If the first result equals 1f8b and the second does not return extra header, then it is likely just a gzip file.

The htslib tool probably does some similar check of the first bytes in the input file, to return a true or false identification.

(Bytes via: https://tools.ietf.org/html/rfc1952#page-5 and http://www.htslib.org/doc/bgzip.html)

If you have a directory of files that are all gzip-formatted and you want to make block-compressed versions, and you are using a bash shell, then you could use a for loop, using the .bgz convention:

$ for in_fn in `ls *.vcf.gz`; do out_fn=${in_fn%.*}.bgz; echo ${out_fn}; gunzip -c ${in_fn} | bgzip > ${out_fn}; tabix -p vcf ${out_fn}; done

If you want to follow the .gz convention, you have to do some extra work:

$ for in_fn in `ls *.vcf.gz`; do tmp_fn=${in_fn%.*}.tmp.gz; echo ${tmp_fn}; gunzip -c ${in_fn} | bgzip > ${tmp_fn}; mv ${tmp_fn} ${in_fn}; tabix -p vcf ${in_fn}; rm ${tmp_fn}; done

Note: This second loop is dangerous, as it will overwrite the original gzip file. I would recommend having backups, writing to a separate directory, or just using the .bgz convention.

After seeing this is required here is how you can use that

for VCF in *.vcf.gz
do
    gunzip -c ${VCF} | bgzip  > ${VCF%gz}bgz
    tabix ${VCF%gz}bgz
done