Do not assign reviewers until the first pipeline is done running

Proposal

This change asks MR authors to hold off assigning reviewers until the first pipeline of their MR has finished running. This should make sure that reviewers are not asked to review code which breaks existing features.

I'm assuming that nobody is actually running the entire test suite on their machines and sometimes it's only after the entire pipeline has finished that you discover additional test failures.

Asking the MR author to hold off assigning a reviewer until the first pipeline is done running could reduce the back and forth based on the test failures.