Interacting design attributes affect the quality of object-oriented software. Our previous studies have developed with individual regression models on understandability and testability. Nonetheless, individual scores do not provide a maintainer with direct information of which classes should be reviewed or refactored initially when there is a shortage of time and resources. This article suggests a multi-criteria model that is explainable and based on understandability and testability indicators and can help to refactor on the class level. The framework maintains the two dimensions as two independent indicators, normalizes the values and incorporates project priorities which are configurable to and generates a ranked list of classes with an explanation which is of metric level. Contrary to an opaque aggregate score, the suggested method will reveal why a particular class is prioritized and which indicators are monitored that allow it to be ranked. The paper establishes the framework architecture, scoring formulation, prioritization algorithm, as well as an empirical validation procedure that relies on object-oriented systems and judging by independent experts. There are no additional experimental findings; the empirical validation and implementation are necessary and only after that, the effectiveness of the framework will become definite.