feat: Sprint 1 — IPA hardening, regression framework, ground-truth review

Track A (Backend):
- Compound word IPA decomposition (schoolbag→school+bag)
- Trailing garbled IPA fragment removal after brackets (R21 fix)
- Regression runner with DB persistence, history endpoints
- Page crop determinism verified with tests

Track B (Frontend):
- OCR Regression dashboard (/ai/ocr-regression)
- Ground Truth Review workflow (/ai/ocr-ground-truth)
  with split-view, confidence highlighting, inline edit,
  batch mark, progress tracking

Track C (Docs):
- OCR-Pipeline.md v5.0 (Steps 5e-5h)
- Regression testing guide
- mkdocs.yml nav update

Track D (Infra):
- TrOCR baseline benchmark script
- run-regression.sh shell script
- Migration 008: regression_runs table

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

admin-lehrer/lib/navigation.ts

@@ -182,6 +182,24 @@ export const navigation: NavCategory[] = [
       // -----------------------------------------------------------------------
       // KI-Anwendungen: Endnutzer-orientierte KI-Module
       // -----------------------------------------------------------------------
+      {
+        id: 'ocr-regression',
+        name: 'OCR Regression',
+        href: '/ai/ocr-regression',
+        description: 'Regressions-Tests & Ground Truth',
+        purpose: 'Regressions-Tests fuer die OCR-Pipeline ausfuehren. Zeigt Pass/Fail pro Ground-Truth Session, Diff-Details und Verlauf vergangener Laeufe.',
+        audience: ['Entwickler', 'QA'],
+        subgroup: 'KI-Werkzeuge',
+      },
+      {
+        id: 'ocr-ground-truth',
+        name: 'Ground Truth Review',
+        href: '/ai/ocr-ground-truth',
+        description: 'Ground Truth pruefen & markieren',
+        purpose: 'Effiziente Massenpruefung von OCR-Sessions. Split-View mit Confidence-Highlighting, Quick-Accept und Batch-Markierung als Ground Truth.',
+        audience: ['Entwickler', 'QA'],
+        subgroup: 'KI-Werkzeuge',
+      },
       {
         id: 'agents',
         name: 'Agent Management',