Issues making Sloth work with Google Managed Prometheus

[MichaelLiermannAtCharles]

We are attempting to use Sloth in conjunction with Google Managed Prometheus, but are experiencing great difficulty with this - SLOs are generated at the Kubernetes level, but it seems they are not being propagated to GMP at all, as the supplied Grafana SLO dashboards do not show them at all and do not find them on a search. There appears to be no guidance for this use case anywhere on the internet. Do you have any information on how to make this work?

[MichaelLiermannAtCharles]

We have made the required changes in the PodMonitor definition for GMP interoperability, and it is successfully created. But again, no communication appears to be taking place with the GMP operator or the Prometheus service behind it.