[nonameiguess]

One of my earlier jobs a decade ago involved doing pipeline development and Jenkins administration for the on-site developer lab on one of the NRO projects, and I inserted a random build failure code snippet to test that pipelines could recover from builds that failed for unpredictable reasons, like a network error rather than anything actually wrong with the build. I had to do this on the real system because we didn't have funds for a staging environment for the dev environment, and naturally I forgot to get rid of it when I was done. So builds randomly failed for years after that before I remembered and fixed it.