fix: localize the primary data-quality evidence summary (live-caught on Unraid)

Live validation on the deployed fix branch caught a real bug: every data-quality
issue's top-of-page "Evidence summary" line rendered the raw, always-English legacy
evidence.summary string unconditionally -- in all three languages -- even though the
backend has been emitting structured, localizable evidence.signals for a while
(app/services/data_quality.py already documented this exact intent). The frontend
side of that conversion was never finished.

- DataQualityIssueDetail.tsx now renders evidence.signals through the operator's
  locale as the primary summary; the raw evidence.summary string is only visible
  inside "Technical details" (via the existing EvidenceDisclosure JSON dump).
- The four DQ-DEMO-* seed rows that anchor the guided demo's scripted scenarios now
  carry real, accurate signals computed at seed time (duplicate-customer's similarity
  score is the actual SequenceMatcher ratio on the seeded names, not invented) instead
  of only a legacy English sentence.
- Rows with no structured signals (generic filler seed data) fall back to the raw
  text rather than showing a blank summary; the one known placeholder string gets its
  own localized rendering so it never displays as English filler either.
- New regression test: the vehicle_status_conflict evidence summary must show
  localized text and must never contain the specific raw English sentence that was
  live-visible before this fix, in all 3 languages.

151 backend tests, Ruff, mypy green; full local Playwright suite green (a couple of
sequential-run-only flakes, both confirmed to pass in isolation and unrelated to this
change).

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

backend/app/seed_loader.py

@@ -4,6 +4,7 @@ import csv
 import uuid
 from dataclasses import dataclass
 from datetime import UTC, date, datetime, timedelta
+from difflib import SequenceMatcher
 from pathlib import Path
 
 from sqlalchemy import delete, insert, update
@@ -108,11 +109,11 @@ def load_seed(db: Session) -> SeedResult:
 
     customer_id_by_ref: dict[str, uuid.UUID] = {}
     customer_rows = []
+    customer_row_by_ref: dict[str, dict] = {}
     for row in _read_csv("customers.csv"):
         cid = uuid.uuid4()
         customer_id_by_ref[row["public_ref"]] = cid
-        customer_rows.append(
+        customer_row = {
-            {
             "id": cid,
             "public_ref": row["public_ref"],
             "first_name": row["first_name"],
@@ -122,7 +123,8 @@ def load_seed(db: Session) -> SeedResult:
             "postal_code": row["postal_code"] or None,
             "city": row["city"] or None,
         }
-        )
+        customer_rows.append(customer_row)
+        customer_row_by_ref[row["public_ref"]] = customer_row
     db.execute(insert(Customer), customer_rows)
     counts["customers"] = len(customer_rows)
     # Second pass for merged_into (self-referencing FK) since target must exist first.
@@ -223,11 +225,42 @@ def load_seed(db: Session) -> SeedResult:
             return "customer", customer_id_by_ref[entity_ref]
         return "vehicle", vehicle_id_by_ref[entity_ref]
 
+    def _seed_signals(public_ref: str, entity_ref: str, related_refs: list[str]) -> list[dict]:
+        # The four named DQ-DEMO-* rows anchor the guided demo's scripted scenarios, so
+        # they carry real, accurate structured signals (not just a legacy English
+        # sentence) -- the frontend renders these as the primary, localized evidence;
+        # see docs/fleet-ops-correction/current-gap-audit.md §6.
+        if public_ref == "DQ-DEMO-DUPLICATE":
+            a = customer_row_by_ref[entity_ref]
+            b = customer_row_by_ref[related_refs[0]]
+            name_a = f"{a['first_name']} {a['last_name']}".strip().lower()
+            name_b = f"{b['first_name']} {b['last_name']}".strip().lower()
+            ratio = SequenceMatcher(None, name_a, name_b).ratio()
+            return [
+                {"code": "duplicate.exact_email"},
+                {"code": "duplicate.exact_phone"},
+                {"code": "duplicate.same_postal_code"},
+                {"code": "duplicate.similar_name", "params": {"score": round(ratio, 2)}},
+            ]
+        if public_ref == "DQ-DEMO-OVERLAP":
+            return [{"code": "overlap.reserved_bookings", "params": {"refs": related_refs}}]
+        if public_ref == "DQ-DEMO-STATUS":
+            return [{"code": "vehicle.booking_conflict"}]
+        if public_ref == "DQ-DEMO-ATTENTION":
+            return [
+                {
+                    "code": "attention.upcoming_booking_missing_inspection",
+                    "params": {"booking_ref": related_refs[0] if related_refs else ""},
+                }
+            ]
+        return []
+
     dq_rows = []
     now = datetime.now(UTC)
     for row in _read_csv("data_quality_issues.csv"):
         entity_type, entity_id = resolve_entity(row["entity_ref"])
         related_ref = row.get("related_ref") or ""
+        related_refs = related_ref.split("|") if related_ref else []
         dq_rows.append(
             {
                 "id": uuid.uuid4(),
@@ -240,7 +273,8 @@ def load_seed(db: Session) -> SeedResult:
                 "evidence_json": {
                     "summary": row["evidence"],
                     "entity_ref": row["entity_ref"],
-                    "related_refs": related_ref.split("|") if related_ref else [],
+                    "related_refs": related_refs,
+                    "signals": _seed_signals(row["public_ref"], row["entity_ref"], related_refs),
                 },
                 "proposed_action_json": {},
                 "detected_at": now,

frontend/e2e/fleet-ops-correction.spec.ts

@@ -350,3 +350,30 @@ test.describe("route matrix (section 11F)", () => {
     });
   }
 });
+
+test.describe("data-quality evidence summary is localized, not raw English (section 6)", () => {
+  // The primary evidence line at the top of every issue's detail page must render the
+  // structured `evidence.signals` in the operator's language; the legacy English
+  // `evidence.summary` string is a technical fallback only, visible solely inside
+  // "Technical details". Live-caught: this line was unconditionally showing raw
+  // English ("vehicle marked available while reserved bookings conflict") in every
+  // language until fixed.
+  const cases: { lang: string; expectedText: RegExp }[] = [
+    { lang: "nl-BE", expectedText: /overlappende reserveringen/i },
+    { lang: "en-GB", expectedText: /overlapping bookings/i },
+    { lang: "fr-BE", expectedText: /chevauchent|chevauchement/i },
+  ];
+
+  for (const { lang, expectedText } of cases) {
+    test(`vehicle_status_conflict evidence is localized (${lang})`, async ({ page, request }) => {
+      await resetDemoData(request);
+      await loginAsOpsManager(page, lang);
+      await page.goto("/data-quality/DQ-DEMO-STATUS");
+
+      const summarySection = page.locator(".record-surface-evidence");
+      await expect(summarySection).toBeVisible();
+      await expect(summarySection).toContainText(expectedText);
+      await expect(summarySection).not.toContainText("vehicle marked available while reserved bookings conflict");
+    });
+  }
+});

frontend/src/i18n/locales/en-GB/quality.json

@@ -101,6 +101,17 @@
         "whyItMatters": "An incorrect status can make an unavailable vehicle appear bookable, or keep an available vehicle hidden from the fleet."
       }
     },
+    "evidence": {
+      "duplicate.exact_email": "Identical email address",
+      "duplicate.exact_phone": "Identical phone number",
+      "duplicate.same_postal_code": "Same postal code",
+      "duplicateSimilarName": "Strongly similar name (score {{score}})",
+      "missingField": "Missing field: {{field}}",
+      "overlapReservedBookings": "Overlapping bookings: {{refs}}",
+      "odometerRegression": "Booking {{laterRef}} recorded {{laterKm}} km, below the {{earlierKm}} km recorded by earlier booking {{earlierRef}}.",
+      "upcomingBookingMissingInspection": "Booking {{bookingRef}} starts soon but the required operational inspection is still missing.",
+      "seedPlaceholder": "Synthetic seed data with no further detail."
+    },
     "duplicateCustomer": {
       "heading": "Compare and merge",
       "description": "Choose the canonical customer and review each conflicting field.",

frontend/src/i18n/locales/fr-BE/quality.json

@@ -101,6 +101,17 @@
         "whyItMatters": "Un statut incorrect peut faire apparaître un véhicule indisponible comme réservable, ou masquer un véhicule disponible de la flotte."
       }
     },
+    "evidence": {
+      "duplicate.exact_email": "Adresse e-mail identique",
+      "duplicate.exact_phone": "Numéro de téléphone identique",
+      "duplicate.same_postal_code": "Même code postal",
+      "duplicateSimilarName": "Nom fortement similaire (score {{score}})",
+      "missingField": "Champ manquant : {{field}}",
+      "overlapReservedBookings": "Réservations qui se chevauchent : {{refs}}",
+      "odometerRegression": "La réservation {{laterRef}} a enregistré {{laterKm}} km, en dessous des {{earlierKm}} km enregistrés par la réservation antérieure {{earlierRef}}.",
+      "upcomingBookingMissingInspection": "La réservation {{bookingRef}} commence bientôt mais l'inspection opérationnelle requise est encore manquante.",
+      "seedPlaceholder": "Donnée de démonstration synthétique sans autre détail."
+    },
     "duplicateCustomer": {
       "heading": "Comparer et fusionner",
       "description": "Choisissez le client de référence et examinez chaque champ divergent.",

frontend/src/i18n/locales/nl-BE/quality.json

@@ -101,6 +101,17 @@
         "whyItMatters": "Een foutieve status kan een niet-beschikbaar voertuig boekbaar laten lijken, of een beschikbaar voertuig verbergen voor het wagenpark."
       }
     },
+    "evidence": {
+      "duplicate.exact_email": "Identiek e-mailadres",
+      "duplicate.exact_phone": "Identiek telefoonnummer",
+      "duplicate.same_postal_code": "Dezelfde postcode",
+      "duplicateSimilarName": "Sterk gelijkende naam (score {{score}})",
+      "missingField": "Ontbrekend veld: {{field}}",
+      "overlapReservedBookings": "Overlappende reserveringen: {{refs}}",
+      "odometerRegression": "Boeking {{laterRef}} registreerde {{laterKm}} km, lager dan de {{earlierKm}} km van eerdere boeking {{earlierRef}}.",
+      "upcomingBookingMissingInspection": "Boeking {{bookingRef}} start binnenkort maar de vereiste operationele inspectie ontbreekt nog.",
+      "seedPlaceholder": "Synthetisch demogegeven zonder verder detail."
+    },
     "duplicateCustomer": {
       "heading": "Vergelijken en samenvoegen",
       "description": "Kies de klant die behouden blijft en bekijk elk afwijkend veld.",

frontend/src/pages/DataQualityIssueDetail.tsx

@@ -46,6 +46,79 @@ function EvidenceDisclosure({ issue }: { issue: IssueDetail }) {
   );
 }
 
+interface EvidenceSignal {
+  code: string;
+  params?: Record<string, unknown>;
+}
+
+// The backend never emits prose for evidence -- only stable signal codes + raw data
+// params (see app/services/data_quality.py::_open_issue). This is the one place that
+// turns them into the operator's selected language; the legacy `evidence.summary`
+// string is a technical fallback only, shown solely inside EvidenceDisclosure above.
+// A handful of seed-only rows (generic filler, not tied to a scripted demo scenario)
+// carry no structured signals -- rather than show nothing, this known placeholder
+// summary gets a localized rendering. Any other un-signalled legacy text still falls
+// back to the raw string (better than a blank primary evidence area), but only
+// "Technical details" is meant to guarantee raw-text visibility.
+const SEED_PLACEHOLDER_SUMMARY = "Synthetic deterministic seed issue";
+
+function EvidenceSignalList({ issue }: { issue: IssueDetail }) {
+  const { t } = useTranslation("quality");
+  const { formatNumber } = useLocaleFormat();
+  const signals = (issue.evidence.signals as EvidenceSignal[] | undefined) ?? [];
+  if (signals.length === 0) {
+    const rawSummary = String(issue.evidence.summary ?? "");
+    if (!rawSummary) return null;
+    const text =
+      rawSummary === SEED_PLACEHOLDER_SUMMARY
+        ? t("detail.evidence.seedPlaceholder")
+        : rawSummary;
+    return (
+      <ul className="evidence-signal-list">
+        <li>{text}</li>
+      </ul>
+    );
+  }
+
+  function describe(signal: EvidenceSignal): string {
+    const { code, params = {} } = signal;
+    if (code.startsWith("vehicle.")) {
+      return t(`detail.vehicleStatusConflict.reasonCodes.${code}`, { defaultValue: code });
+    }
+    switch (code) {
+      case "missing_field":
+        return t("detail.evidence.missingField", {
+          field: t(`detail.missingField.fields.${params.field}`, { defaultValue: String(params.field) }),
+        });
+      case "overlap.reserved_bookings":
+        return t("detail.evidence.overlapReservedBookings", {
+          refs: Array.isArray(params.refs) ? params.refs.join(" ↔ ") : "",
+        });
+      case "odometer.regression":
+        return t("detail.evidence.odometerRegression", {
+          laterRef: params.later_ref,
+          laterKm: formatNumber(Number(params.later_km ?? 0)),
+          earlierRef: params.earlier_ref,
+          earlierKm: formatNumber(Number(params.earlier_km ?? 0)),
+        });
+      case "duplicate.similar_name":
+        return t("detail.evidence.duplicateSimilarName", { score: params.score });
+      case "attention.upcoming_booking_missing_inspection":
+        return t("detail.evidence.upcomingBookingMissingInspection", { bookingRef: params.booking_ref });
+      default:
+        return t(`detail.evidence.${code}`, { defaultValue: code });
+    }
+  }
+
+  return (
+    <ul className="evidence-signal-list">
+      {signals.map((signal, index) => (
+        <li key={`${signal.code}-${index}`}>{describe(signal)}</li>
+      ))}
+    </ul>
+  );
+}
+
 function DuplicateCustomerPanel({ issue, onResolved }: { issue: IssueDetail; onResolved: () => void }) {
   const { t } = useTranslation("quality");
   const { user } = useAuth();
@@ -726,8 +799,11 @@ export function DataQualityIssueDetail() {
       <section className="record-surface" aria-label={t("detail.summary.rule")}><dl className="detail-grid">
         <div><dt>{t("detail.summary.rule")}</dt><dd>{t(`ruleTypes.${issue.rule_type}`, { defaultValue: issue.rule_type.replace(/_/g, " ") })}</dd></div>
         <div><dt>{t("detail.summary.entity")}</dt><dd>{issue.entity_type === "vehicle" ? <Link to={`/vehicles/${issue.entity_ref}`}>{issue.entity_ref}</Link> : issue.entity_ref}</dd></div>
-        <div><dt>{t("detail.summary.evidenceSummary")}</dt><dd>{String(issue.evidence.summary ?? "")}</dd></div>
       </dl>
+      <div className="record-surface-evidence">
+        <dt>{t("detail.summary.evidenceSummary")}</dt>
+        <EvidenceSignalList issue={issue} />
+      </div>
       <EvidenceDisclosure issue={issue} /></section>
 
       <RuleExplainer ruleType={issue.rule_type} />

frontend/src/styles.css

@@ -239,6 +239,10 @@ details summary { cursor: pointer; color: var(--teal-dark); }.data-table details
 .tabs button { position: relative; min-height: 44px; padding: 8px 13px; border: 0; background: transparent; color: var(--muted); font-size: .72rem; font-weight: 700; text-transform: capitalize; cursor: pointer; }
 .tabs button.active { color: var(--teal-dark); }.tabs button.active::after { content: ""; position: absolute; inset: auto 7px -1px; height: 2px; background: var(--teal); }
 .record-surface { padding: 18px; margin-bottom: 18px; }
+.record-surface-evidence { padding: 0 18px 16px; }
+.record-surface-evidence dt { margin: 0 0 6px; color: var(--muted); font-size: .63rem; font-weight: 700; text-transform: uppercase; letter-spacing: .065em; }
+.record-surface-evidence .evidence-signal-list { margin: 0; padding-left: 18px; color: var(--ink); font-size: .82rem; font-weight: 700; line-height: 1.6; }
+.record-surface-evidence .evidence-signal-list li { margin-bottom: 3px; }
 .about-card h2 { margin: 0 0 8px; color: var(--ink); font-size: 1rem; letter-spacing: -.015em; }
 .about-card p { margin: 0; color: var(--muted); font-size: .82rem; line-height: 1.62; }
 .about-card p code { padding: 1px 5px; background: var(--surface-subtle); border-radius: 4px; font-size: .78rem; }