Stephen Benjamin began a research project into sippy scraping prow directly, instead of via testgrid, originally as a side project to try to investigate costs to pull request CI runs that are unnecessarily re-run due to flaky jobs and infrastructure. We now see potential for this effort to actually replace how sippy currently works today in scraping testgrid. Prow exposes more fine grained data we cannot infer from testgrid such as test suite names, and the build cluster where we're running.

We hope to pursue this effort and leverage the new data to help our automation in TRT-333.