Client-side pre-flight lint on publish silently soft-fails with HTTP 400 instead of erroring

Reproduction steps:

1. Run railcall market publish . --type=module --price=0 on a module ready to publish.
2. Observe the output during the pre-flight lint step.

Expected: A pre-flight check that can't reach its endpoint should either error clearly
(so the operator knows a check was skipped) or succeed silently if genuinely not needed
— not print an HTTP error and continue as if nothing happened.

Actual: Output showed "Skipping pre-flight lint: HTTP Error 400: Bad Request" followed
by "Server-side quality gate still runs on publish." The publish itself succeeded and
the resulting listing was complete and correct, so the server-side gate did catch
anything that mattered here — but the message implies the client-side check is quietly
degrading against a broken/missing endpoint rather than failing loudly, which could mask
a real problem in a future case where the server-side gate doesn't catch everything the
client-side one would have.

Suggested fix: either fix the endpoint the client-side lint hits, or make the failure
more visible (a warning-level message rather than a quiet skip) so operators know a
check didn't run.

Station version (railcall version): station-v1.5.8
Module slug + version: muhammad-akif-janjua/slack-guard v0.4.1