Given the user can soon have multiple config sources for CI, we now store
what type at the time of the pipeline run we chose. This will give us
insight into what triggered the new pipeline so we can display it to the
enduser.