The modern way is to simply invoke use Test::More; at the top, and at the end have done_testing().

If there were problems, they would generate error messages, which are detected by TAP; if there are no errors, the system counts up the successful tests and uses that in the final message.