In many of our test plans and demonstration documents, we just refer to the running of specific test scripts (i.e. sanity_dne.sh). This does not help the reader understand what is actually being tested by the test script unless the reader wants to read the bash script's comments and code to determine what is being tested.

We should have a single, maintained description of what is tested by each test script. Preferably this description is versioned so we can refer to a version of the script's description from test plans and demonstration documents.