Library(prolog_coverage) - working with predicates generated during source code load

Help!
1

$ swipl --version
SWI-Prolog version 10.0.0 for x64-win64

Using library(prolog_coverage) on a module that uses library(persistency).

For code coverage for persistent/1 was able to increase the code coverage by adding tests for the dynamicly created four predicates resulting in this

   90 ### ⮰
      ++20 ⮰
      ++4,727 ⮰
      ++21,486 ⮰
      +55,950-20             :- persistent(pred_def(name:atom, arity:integer, module:atom, file:atom, line:integer)).

There is now just the one remaining ### that I have no idea of what it could be.

Any ideas? Is it a predicate, an artifact of code coverage that can be removed, other?

Any ideas on how to get a line in the cov report to show what it could be?

Gave Claude the links to the SWI-Prolog library code on GitHub and noted the relevant files in the project and Claude thinks

   90 ### ⮰                          ← The directive itself: never "executed"
      ++20 ⮰                          ← Generated predicate variant 1
      ++4,727 ⮰                       ← Generated predicate variant 2
      ++21,486 ⮰                      ← Generated predicate variant 3
      +55,950-20                      ← Generated predicate variant 4
             :- persistent(pred_def(name:atom, arity:integer, module:atom, file:atom, line:i

If that is true would like to find a way to remove such lines from the cov report so that the report shows 100% coverage for the module. Currently Claude will search for ### lines and thus it is like having a wrench thrown into the clock works.

2

Using Claude Code for VS Code was able to modify a local copy of

See this reply for details: