I expect my developers to troubleshoot a failing pipeline too. The pipeline is part of the "ops platform" they're expected to work with.