The nightly rollup and backup say by their exit status whether they managed, and the deploy leaves the cron file off a host the hall schedules

On the shared host the hall's hall-nightly.sh (game-kit, gamehall/ops) runs
both scripts for every game and checks in with Sentry Crons once for all,
so this game's own check-in files are absent there and its exit status is
the word. A box of its own keeps its cron and check-ins as before.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

deploy/wizwar-rollup.sh

@@ -9,7 +9,9 @@
 # one count per address per day.
 #   Cron: see SCHEDULE below and /etc/cron.d/wizwar-rollup (deploy.sh installs both).
 #   Sentry Crons check-in: /root/.wizwar-sentry-cron-rollup holds the URL
-#   (absent = no check-ins).
+#   (absent = no check-ins). On a host the hall shares, hall-nightly.sh
+#   runs this and checks in once for every game, so the file is absent and
+#   the exit status is the word.
 set -u
 OUT="/var/lib/wizwar/rollup.jsonl"
 LOG="/var/log/wizwar/rollup.log"
@@ -148,4 +150,4 @@ tmp = out + ".tmp"
 open(tmp, "w").write("\n".join(rows) + "\n"); os.replace(tmp, out)
 print("%s rollup %s: %d requests, %d human addresses, %d new rooms, %d human seats" % (datetime.datetime.now(datetime.timezone.utc).isoformat(), day, req, len(hips), new_rooms, sum(seats.values())))
 PY
-then checkin ok; else checkin error; fi
+then checkin ok; else checkin error; exit 1; fi