Skip if the post is no longer status=publishing when the job runs, so
a stray duplicate (e.g. clicking Publish twice while no worker was
running) can't publish the same post to WordPress a second time.
Also root-caused a stuck-in-"publishing" post reported live: the
queue worker started manually last session was a bare background
process, not a real service, and had died. Replaced it with a
systemd unit (muninn-queue-worker.service, enabled + running) so
the Publish button keeps working across reboots/restarts.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
