feat(notifications): move Notification Routing to Skipper tier (#1145)

* feat(notifications): move Notification Routing to Skipper tier

Notification routing is automation (route alerts to channels by rules),
not enterprise compliance. Aligning the gate with Skipper makes the tier
boundary read consistently with the rest of the automation surface
(webhooks, auto-update, auto-heal, scheduled tasks).

Backend: requireAdmiral -> requirePaid on the five /api/notification-routes
endpoints. Dashboard configuration-status now exposes the routing-rules
row to any paid tier.

Frontend: settings registry tier flipped to skipper; the Admiral wrapper
around NotificationRoutingSection is removed (the inner CapabilityGate
stays, preserving forward-compat with older remote nodes).

Tests: added a tier-enforcement describe block covering Skipper (200) and
Community (403 PAID_REQUIRED on all five endpoints).

Docs: refreshed alerts-notifications, licensing, overview, dashboard,
troubleshooting, and reference/settings; cleaned one fence-spec line per
Directive 31.

* fix(notifications): address audit findings on tier-move PR

Docs: rewrite three lines that survived the initial sweep. The dashboard
"you do not see a locked placeholder" clause and the settings.mdx
"hidden on Community and Skipper" phrase were Directive 31 fence-spec.
The alerts-notifications troubleshooting note still said "an Admiral
routing rule" and contradicted the tier move.

Tests: the Community-negative cases on POST/PUT/DELETE/POST :id/test
could not distinguish requirePaid from a stray requireAdmiral, because
Community fails on the tier check before variant is read. Adding
Skipper-positive coverage per endpoint locks the gate identity in.
Replace the leaky mockReturnValueOnce with a per-test mockReturnValue
plus an afterEach restore so spies cannot bleed across tests.

docs/features/licensing.mdx

@@ -41,6 +41,7 @@ For larger deployments, an **Enterprise** tier is available with custom pricing,
 
 - Fleet View with search, sort, filter, and node-card drill-down
 - Webhooks (incoming, to trigger deploys from CI/CD)
+- Notification routing (per-stack and per-category rules to Discord, Slack, or any webhook)
 - Atomic deployments with rollback
 - Auto-update policies for stack images
 - Auto-heal policies
@@ -61,7 +62,6 @@ For larger deployments, an **Enterprise** tier is available with custom pricing,
 - Host Console (a browser-based terminal on the Sencho host)
 - API tokens for CI/CD scripts
 - Private and custom registry credentials
-- Notification routing
 - Sencho Mesh (cross-node container networking)
 - Sencho Cloud Backup
 - Auto-update of the managed Trivy binary