Independent pytest runtime reproduction trial (supplemental receipts)

[shleder]

I’m exploring a small, independent consumer workflow for pytest runtime/config reproductions with ReproProof. For one ordinary API- or CLI-triggered pytest run—such as an interpreter, dependency, or config mismatch—I’d like to compare the receipt with the normal minimal reproduction and hear whether it adds useful context. The safe flow is reproproof capture in a disposable fixture, then reproproof verify receipt.json --cwd ./project --no-replay --json first; the receipt is supplemental, and I would share only the Markdown summary plus feedback issue #18, never raw JSON, secrets, or local paths. If a pytest user or maintainer is open to a small independent trial, I’d appreciate a note here.