feat(trackers): user filters for Gitea, webhook polling cleanup, dashboard navigability

- Gitea: NotificationTracker now exposes sender allowlist / blocklist filters
  via MultiEntitySelect, populated from Gitea /users/search merged with past
  EventLog senders so the picker is useful before the first webhook arrives.
- Webhook providers (gitea, planka, webhook): stop scheduling interval polling
  jobs on tracker create/update/startup; hide the "every Xs" indicator in the
  tracker list since there is no polling.
- Dashboard: stat cards are now <a> links that route to providers, trackers,
  targets, command-trackers, or scroll to the events panel. Provider deck
  rows highlight the target provider on click.
- Command trackers / command configs: auto-reselect the right config when the
  provider type changes (matches notification-tracker behavior).
- Migration: drop legacy batch_duration column from notification_tracker —
  the field is gone from the model but its NOT NULL constraint blocked
  inserts on older DBs.
- Docs: refresh entity-relationships.md with current NotificationTracker
  fields (filters, adaptive_max_skip, default_*_config_id).

packages/server/src/notify_bridge_server/services/scheduler.py

@@ -378,6 +378,8 @@ async def _load_tracker_jobs() -> None:
 
     tz = await _load_app_timezone()
 
+    from notify_bridge_core.providers.capabilities import get_capabilities
+
     for tracker in trackers:
         job_id = f"tracker_{tracker.id}"
         if scheduler.get_job(job_id):
@@ -386,6 +388,18 @@ async def _load_tracker_jobs() -> None:
         ptype = provider_types.get(tracker.provider_id, "")
         filters = tracker.filters or {}
 
+        # Webhook-based providers receive events via inbound HTTP — there is
+        # nothing to poll. Scheduling an interval job for them just wakes up
+        # check_tracker every scan_interval seconds to immediately return,
+        # wasting CPU and DB queries for no work.
+        caps = get_capabilities(ptype) if ptype else None
+        if caps and caps.webhook_based:
+            _LOGGER.debug(
+                "Skipping interval scheduling for webhook tracker %d (%s, type=%s)",
+                tracker.id, tracker.name, ptype,
+            )
+            continue
+
         # Scheduler providers can use cron triggers
         if ptype == "scheduler" and filters.get("schedule_type") == "cron":
             cron_expr = filters.get("cron_expression", "")
@@ -450,6 +464,29 @@ def _add_cron_job(
     )
 
 
+async def _is_webhook_tracker(tracker_id: int) -> bool:
+    """Return True iff the tracker's provider type is webhook-based.
+
+    Looks up provider type once via the capabilities registry. Used by
+    ``schedule_tracker`` to short-circuit interval scheduling.
+    """
+    from sqlmodel import select
+    from sqlmodel.ext.asyncio.session import AsyncSession
+    from notify_bridge_core.providers.capabilities import get_capabilities
+    from ..database.engine import get_engine
+    from ..database.models import NotificationTracker, ServiceProvider as ServiceProviderModel
+
+    async with AsyncSession(get_engine()) as session:
+        tracker = await session.get(NotificationTracker, tracker_id)
+        if tracker is None:
+            return False
+        provider = await session.get(ServiceProviderModel, tracker.provider_id)
+        if provider is None:
+            return False
+        caps = get_capabilities(provider.type)
+        return bool(caps and caps.webhook_based)
+
+
 async def schedule_tracker(
     tracker_id: int,
     interval: int,
@@ -461,6 +498,10 @@ async def schedule_tracker(
     ``adaptive_max_skip`` mirrors the DB column and is registered with the
     adaptive module-state so tick-time skip decisions don't re-query the DB.
     Pass ``None`` or ``0`` to disable back-off for the tracker.
+
+    Webhook-based providers receive events via inbound HTTP and have nothing
+    to poll, so this no-ops for them — preventing scan_interval from creating
+    useless wakeups via the API create/update path.
     """
     scheduler = get_scheduler()
     job_id = f"tracker_{tracker_id}"
@@ -474,6 +515,13 @@ async def schedule_tracker(
     if scheduler.get_job(job_id):
         scheduler.remove_job(job_id)
 
+    # Webhook-based providers don't poll — skip job creation entirely.
+    if await _is_webhook_tracker(tracker_id):
+        _LOGGER.debug(
+            "Skipping interval scheduling for webhook tracker %d", tracker_id,
+        )
+        return
+
     if cron_expression:
         try:
             tz = await _load_app_timezone()