A "build" often includes running the unit test suite to verify that the software works in the build environment, so randomness in the test suite can indeed lead to non-reproducible output: when a bad RNG seed causes a test failure and thus a failed build. I've spent my fair share of time debugging issues like this.

Use a PRNG seeded with the hash of the build manifest.

I don't know what a build manifest is but yeah, that sounds great. However, the test suite needs to support this. Plenty of times they do not.