Add queue-based gradual discovery with shuffled call ordering and progress UI

Each yt-dlp call is now an independent task (one search query, one trending
fetch, one graph channel fetch). Tasks are shuffled together so we don't fire
10 searches in a row, then enqueued with 30-90s random gaps between them —
a full sweep of ~17 tasks completes in roughly 10-25 minutes instead of
hammering YouTube with 21 calls back-to-back.

Fast signals (community, category clusters) still run synchronously at
schedule time since they're pure SQL.

Progress is tracked per-user (total/done/running) and exposed on
GET /api/discovery/status. The Discovery page polls every 10s while
running and shows a progress bar + "Finding channels… X / Y" in the header.
The auto-discovery daemon skips scheduling if a manual sweep is already running.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

backend/main.py

@@ -175,9 +175,11 @@ def on_startup():
     finally:
         db.close()
 
-    # Backfill descriptions for videos that don't have them yet (runs in background)
+    # Start discovery worker and backfill enrichment
     import threading
     from .routers.channels import _enrich_missing_task, _index_channels_batch
+    from .services.discovery import start_discovery_worker
+    start_discovery_worker()
     threading.Thread(target=_enrich_missing_task, args=(50,), daemon=True).start()
 
     def _auto_sync_daemon():
@@ -222,7 +224,7 @@ def on_startup():
         import time as _time
         from datetime import datetime as _dt, timedelta as _td
         from sqlalchemy import text as _text
-        from .services.discovery import run_full_discovery
+        from .services.discovery import schedule_discovery, get_discovery_progress
 
         # Wait 5 minutes after startup before the first check so the app can
         # finish initialising and existing enrichment tasks can settle.
@@ -244,13 +246,12 @@ def on_startup():
                         last = row["last_discovery_run"]
                         if last is None or (_dt.utcnow() - _dt.fromisoformat(str(last))) > _td(hours=23):
                             uid = row["user_id"]
+                            # Skip if a manual sweep is already running
+                            prog = get_discovery_progress(uid)
+                            if prog and prog.get("running"):
+                                continue
                             regions = [r.strip().upper() for r in (row["discovery_regions"] or "US,SE").split(",") if r.strip()]
-                            run_full_discovery(db, uid, regions)
-                            db.execute(
-                                _text("UPDATE user_settings SET last_discovery_run = :now WHERE user_id = :uid"),
-                                {"now": _dt.utcnow(), "uid": uid},
-                            )
-                            db.commit()
+                            schedule_discovery(uid, regions)
                 finally:
                     db.close()
             except Exception: