feat(search): full-platform search overhaul + view tracking + notes bucket

Service rewrite covers 14 entity buckets (clients, residential clients,
yachts, companies, interests, residential interests, berths, invoices,
expenses, documents, files, reminders, brochures, tags, notes, navigation)
with prefix tsquery + trigram fallback, phone-digit normalization,
and JOINs to client_contacts for email matching.

New `notes` bucket searches across the four note tables (client,
interest, yacht, company) via UNION + parent-entity label resolution
(berth mooring for interests, name for yachts/companies). Renders at
the bottom of the dropdown so broad-content matches don't crowd
entity-specific hits — per the user's "low-noise" preference.

Recently-viewed tracking persists last 20 entity views per user in
Redis sorted set; CommandSearch surfaces them as the dropdown's
default state and applies affinity ranking when the user types.

ID-resolve endpoint accepts pasted UUIDs (or invoice numbers like
`INV-2025-001`) and routes the rep straight to the entity, skipping
the normal search bucket.

Audit search service gains `entityIds[]` array filter for the new
loadClientActivityAggregated() path.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src/components/search/highlight-match.tsx

@@ -0,0 +1,54 @@
+import { Fragment, type ReactNode } from 'react';
+
+/**
+ * Wrap occurrences of `query` inside `text` with `<mark>` so the user
+ * can see why a result matched. Case-insensitive, escapes regex meta-
+ * chars in the query so a paste of "INV-2025" doesn't blow up.
+ *
+ * Tokenized matching — splits the query on whitespace and highlights
+ * each token independently, so "joh smi" highlights both "Joh" and
+ * "Smi" in "John Smith". Mirrors the prefix-tsquery the server uses.
+ */
+export function HighlightMatch({
+  text,
+  query,
+  className,
+}: {
+  text: string | null | undefined;
+  query: string;
+  className?: string;
+}): ReactNode {
+  if (!text) return null;
+  const tokens = query
+    .trim()
+    .split(/\s+/)
+    .filter((t) => t.length > 0)
+    .map(escapeRegex);
+  if (tokens.length === 0) return text;
+
+  const re = new RegExp(`(${tokens.join('|')})`, 'gi');
+  const parts = text.split(re);
+
+  return (
+    <span className={className}>
+      {parts.map((part, i) => {
+        if (i % 2 === 1) {
+          // The capture group lands on odd indices.
+          return (
+            <mark
+              key={i}
+              className="bg-brand/15 text-foreground rounded-[2px] px-[1px] font-medium"
+            >
+              {part}
+            </mark>
+          );
+        }
+        return <Fragment key={i}>{part}</Fragment>;
+      })}
+    </span>
+  );
+}
+
+function escapeRegex(input: string): string {
+  return input.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+}

src/components/search/track-entity-view.tsx

@@ -0,0 +1,33 @@
+'use client';
+
+import { useTrackEntityView } from '@/hooks/use-track-entity-view';
+
+/**
+ * Render-only client component that records "the user opened this
+ * entity" via the global-search recently-viewed Redis log. Drop into a
+ * server-rendered detail page as `<TrackEntityView type="client" id={…} />`
+ * — the component renders nothing.
+ *
+ * Centralises the tracking call so future changes (debounce, schema,
+ * batching) only need to touch one file rather than every detail page.
+ */
+export function TrackEntityView({
+  type,
+  id,
+}: {
+  type:
+    | 'client'
+    | 'residential-client'
+    | 'yacht'
+    | 'company'
+    | 'interest'
+    | 'residential-interest'
+    | 'berth'
+    | 'invoice'
+    | 'expense'
+    | 'document';
+  id: string | null | undefined;
+}) {
+  useTrackEntityView(type, id);
+  return null;
+}