This process is described in the attached diagram.

By determining what new packages were successfully built in a metro stage build run, we can then attach relevant test execution results to issues in Jira, allowing us to get feedback on whether a particular catpkg is building properly.

To get this working well, we would likely need to use a "testing" branch of kit-fixups in which we would commit our more experimental packages that are in need of build testing. So we may want to look at the "future work" of auto-promotion of changes into master sooner.

It would also be possible to use feature branches for each merge, though I think that could become cumbersome. Potentially, each developer would have their own testing branch which could get independently tested using this method, potentially by doing a metro run on their own system.