To automate the task of testing an application it is possible with Trinc-Prolog to create test scripts. In test script files the goals to prove and all steps for proving the goal of Trinc-Prolog are stored. Test scripts can then be replayed at a later time to check if the input still produces the expected output. The input and the expected output are both stored in a single test script file. It is possible to record mouse and keyboard events to automate testing of applications that have a graphical user interface.
A test can be replayed as a blackbox or a whitebox test. When a test is replayed as a blackbox test only the user input and solutions of goals are compared. This makes it possible to check if a new version of an algorithm still produces correct results with a test script of an older version of the same algorithm. While replaying a whitebox test the user input and all the steps of the Trinc-Prolog inference engine and the output are compared.