Skip to main content

baseline_cannot_run

Function baseline_cannot_run 

Source
pub fn baseline_cannot_run(results: &[CheckResult]) -> Vec<String>
Expand description

Checks whose COMMAND could not run at all, by name.

A check that fails because the code is broken is the point — that is the red half of the red-green baseline. A check that fails because its program does not exist is a different thing wearing the same exit status: the contract can never go green no matter what the session writes, so the run is doomed before the first edit and will spend its whole iteration budget discovering that.

Found on a live run: derivation produced python -m pytest ... on a machine with only python3, and twelve iterations of real inference went into a contract that was unsatisfiable from the start. Nothing distinguished it from an ordinary red baseline.

Exit 127 is the shell’s “command not found”; None means the runtime could not spawn it at all. Both mean the same to a caller: this check is not a test of anything yet.