feat(goldset): interactive gold-set tagging page (#81.7/#81.8)

Replaces the CSV-edit workflow with an in-app tagging page so the chair/Dafna
can label the extraction-quality gold-set by clicking, and see validator
precision/recall live.

Schema (V29): halacha_goldset — a stratified, human-tagged evaluation batch
(is_holding / correct_type / quote_complete, NULL until tagged).

db.py:
- goldset_create_sample (stratified round-robin over case×rule_type, idempotent),
- goldset_list (items + halacha content + the machine's own labels),
- goldset_tag (partial — one field at a time for keyboard tagging),
- goldset_score (ports the script's P/R/F1: each validator scored as a
  not-a-holding detector against the human tags — the #81.8 input).

API: GET /api/goldset, POST /api/goldset/sample, GET /api/goldset/score,
PATCH /api/goldset/{id}.

web-ui:
- lib/api/goldset.ts (hooks),
- components/goldset/goldset-panel.tsx — card-per-item, keyboard-first
  (J/K nav, H/N holding, C/X quote), progress bar, hide-tagged toggle, and a
  collapsible live score table,
- app/goldset/page.tsx + nav link "מדגם-זהב" under ידע ולמידה.

Methodology guard kept explicit in UI + docstrings: tags are HUMAN ground truth,
no AI pre-fill (circular bias). Populated a 150-item stratified batch.

Verified: backend create/list/tag/score against the live DB; tsc --noEmit 0;
py_compile ok. (Local Turbopack build blocked by worktree symlink — CI builds clean.)

Invariants: G1 (eval set modeled at source in its own table); G2 (reuses the same
halacha_quality validators the extractor runs — no parallel scoring logic).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

web-ui/src/app/goldset/page.tsx

@@ -0,0 +1,41 @@
+"use client";
+
+import Link from "next/link";
+import { AppShell } from "@/components/app-shell";
+import { GoldsetPanel } from "@/components/goldset/goldset-panel";
+
+/**
+ * Gold-set tagging page (#81.7 / #81.8).
+ *
+ * Interactive review of a stratified halacha sample. The chair/Dafna labels each
+ * item (is_holding / correct_type / quote_complete); those human labels are the
+ * ground truth that measures the extraction validators and recalibrates the
+ * auto-approve threshold. Tags MUST be human — no AI pre-fill (circular bias).
+ */
+export default function GoldsetPage() {
+  return (
+    <AppShell>
+      <section className="space-y-6">
+        <header>
+          <nav className="text-[0.78rem] text-ink-muted mb-1">
+            <Link href="/" className="hover:text-gold-deep">בית</Link>
+            <span aria-hidden> · </span>
+            <span className="text-navy">מדגם-זהב לתיוג</span>
+          </nav>
+          <h1 className="text-navy mb-0">מדגם-זהב לתיוג איכות</h1>
+          <p className="text-ink-muted text-sm mt-1 max-w-3xl">
+            מדגם מרובד של הלכות שחולצו. לכל הלכה הכריעו שלוש שאלות —
+            <strong> האם זו הלכה אמיתית</strong>, <strong>מה הסוג הנכון</strong>,
+            ו<strong>האם הציטוט שלם</strong>. ההכרעות שלכם הן אמת-המידה שמודדת את
+            דיוק המחלץ ומכיילת את סף-האישור האוטומטי. שיפוט משפטי אנושי בלבד —
+            לא תיוג-AI (כדי למנוע הטיה מעגלית).
+          </p>
+        </header>
+
+        <div className="h-[2px] bg-gradient-to-l from-transparent via-gold to-transparent" />
+
+        <GoldsetPanel />
+      </section>
+    </AppShell>
+  );
+}