fix scheduler misfire: daily jobs silently skipped on a busy event loop

AsyncIOScheduler was constructed with no job_defaults, so APScheduler's default
misfire_grace_time of 1s applied. In this single-process app the scheduler shares
one event loop with the API and all other jobs, so when a daily job came due
while the loop was busy (e.g. the scanner mid-run), the fire was processed >1s
late, flagged a misfire, and skipped — while next_run still advanced 24h, making
the job look healthy though it never ran. Set a generous grace window (1h),
coalesce missed runs into a single catch-up, and cap concurrency at 1.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

app/scheduler.py

@@ -41,8 +41,22 @@ from app.services.ticker_universe_service import bootstrap_universe
 
 logger = logging.getLogger(__name__)
 
-# Module-level scheduler instance
+# Module-level scheduler instance.
-scheduler = AsyncIOScheduler()
+#
+# job_defaults matter a lot here: this is a single-process app, so the scheduler
+# shares one event loop with the API and every other job. APScheduler's default
+# misfire_grace_time is just 1 second — if the loop is busy at the instant a
+# daily job is due (e.g. the scanner is mid-run), the fire is processed late,
+# flagged a misfire, and SILENTLY SKIPPED while next_run still advances 24h. So
+# we grant a generous grace window, coalesce missed runs into one catch-up, and
+# cap each job at a single concurrent instance.
+scheduler = AsyncIOScheduler(
+    job_defaults={
+        "coalesce": True,
+        "max_instances": 1,
+        "misfire_grace_time": 3600,  # tolerate a busy loop; a daily job up to 1h late is fine
+    }
+)
 
 # Track last successful ticker per job for rate-limit resume
 _last_successful: dict[str, str | None] = {