Testing frequently changing api, scenario? please refer the details below :-

I am working on a custom test framework (in java), There are 4 components in this:-

1. API to Tests : These are the classes, functions etc.. which are under test.

Example: 
ClassAPI1 {apiFunction11, apiFunction12, apiFunction13, apiFunction14}
ClassAPI2 {apiFunction21, apiFunction22, apiFunction23, apiFunction24}

It is bundled in a jar, lets say : api.jar

2. Library : There are reusable helper functions, utility methods , logging functions, assertion functions etc.

Example:
ClassLib1 { libFunction1, libFunction2, libFunction3, libFunction4 } 
ClassLib2 { libFunction1, libFunction2, libFunction3, libFunction4 }

3. Tests : There are tests which are calling api’s + Library in specific order , so that it become a use case or business scenario.

Example :

Tests1Class {

    @Metadata (Handled by Test Runner Framework) like test owner, testid etc...
    ClassAPI1.apiFunction11
    ClassAPI2.apiFunction21
    ClassAPI1.apiFunction13
    ClassLib1.libFunction2
    ClassLib2.libFunction4
    
    }  // End of test case.

Tests and Library also bundled in a jar lets say : tests-lib.jar

4. Test Runner Framework : It actually read test or list of test cases , execute them and log results. It is separate jar over which user has no control lets say : runner.jar

Now I am expecting frequent changes in API under test and in that case I may have to change my lib + test cases every time if there are change(s) in API’s under test.

My Objective is to have something additional may be a virtual layer, which I don't know as of now how it will look like.

I wish this virtual layer to accommodate these changes so that I don’t have to make changes in lib and my tests or minimal changes when Underlying API is changing.

Can someone suggest any good solution to address this problem?