Maybe a better example would be a setup script. The setup script does not change test anything. After it finishes running it is reported as a successful test. This inflates my report of how many tests were performed. That in turn skews the report of what percentage of the planned tests failed.
Similarly, some scripts perform multiple tests. In such a case the log shows that only one test has passed or failed, when really multiple tests have passed or failed.
Scripts are not tests. TC's log system reports script failures as test failures.
