Merge branch 'main' of ssh://coolify.meghsakha.com:22222/Benjamin_Boenisch/breakpilot-core

2026-05-12 17:45:50 +02:00
parent fc4d5d8c56 f5d4e3bd95
commit edac3aca6c
11 changed files with 407 additions and 210 deletions
@@ -1,194 +1,117 @@
-# Session-Instruktionen: Master Control Qualitaet + Regulation-Source Split
+# Session-Handover: MC Quality + Gap-Analyse + RAG Ingestion

-**Datum:** 2026-05-06
-**Fuer:** Naechste Claude-Session
-**Repo:** breakpilot-core (~/Projekte/breakpilot-core)
+**Datum:** 2026-05-07 bis 2026-05-11 (5 Tage Marathon)
+**Repo:** breakpilot-core + breakpilot-compliance

 ---

-## NAECHSTER SCHRITT: 25 grosse Master Controls aufsplitten
+## ERLEDIGT

-### Problem
+### Master Control Quality Overhaul (Core)
+- **74.5% → 92.8% Accuracy** (13.588 MCs, 83.073 Members)
+- Phase 0: Quality Audit mit Claude Sonnet ($3)
+- Phase 1: Ontologie 31 → 74 Tokens + LLM-Prompt fix
+- Phase 2: 174K Controls re-klassifiziert via Haiku (10 Batches, ~$50)
+- Phase 2b: Generic Tokens gefixt (documentation/procedure → echte Themen, $7.54)
+- Phase 2c: L2 Sub-Topics (2 Runden, 172K Controls, ~$32)
+- Phase 2d: Bad Subtopics gefixt (stakeholder_*, $0.50)
+- Phase 3: Re-Clustering K=18704
+- Phase 4: gpre2 Direct MC (13.588 MCs)
+- Phase 6: Golden Dataset (20 Controls) + 8 Quality Tests (alle grün)
+- **Production Sync:** MCs + Members + Hints + doc_check_controls

-25 Master Controls sind zu generisch (>200 Atomic Controls pro MC). Sie basieren auf generischen Security-Domain-Keywords wie "monitoring", "encryption", "personal_data". Embedding-Clustering allein reicht nicht — die Controls handeln zwar alle von "monitoring", aber fuer unterschiedliche Regulierungen (DSGVO, NIS2, NIST, BSI etc.).
+### doc_check_controls (Core → Production)
+- **1.874 Controls** über 8 Dokumenttypen (DSE, Cookie, Impressum, AGB, Widerruf, DSFA, AVV, Löschkonzept)
+- Jeder mit check_question + pass_criteria + fail_criteria
+- Tabelle `compliance.doc_check_controls` lokal + Production

-### Die 25 betroffenen MCs
+### RAG Ingestion (Core)
+- **126 BAuA PDFs** (TRBS/TRGS/ASR): 27.664 Chunks → `bp_compliance_ce`
+- **OSHA Technical Manual** (23 Kapitel): 7.241 Chunks → `bp_compliance_ce`
+- **OSHA 1910 Subpart O** (Volltext): 745 Chunks
+- **EuGH C-588/21 P**: 216 Chunks
+- **EU 2018/1725**: 842 Chunks → `bp_compliance`
+- **CE-Obligations extrahiert:** 6.141 Obligations → `/tmp/ce_obligations_v2.json`
+- Playwright-Crawler für BAuA + OSHA gebaut

-| MC-ID | Name | Controls | Problem |
-|-------|------|----------|---------|
-| MC-8292 | monitoring | 6.157 | Alles von Video bis Vulnerability |
-| MC-2260 | procedure | 4.176 | Generisch |
-| MC-8302 | alerting | 3.126 | Meldepflichten aller Gesetze gemischt |
-| MC-8306 | personal_data | 3.057 | DSGVO + NIS2 + AT/CH gemischt |
-| MC-8312 | training | 2.572 | |
-| MC-7932 | certificate_management | 2.350 | |
-| MC-8317 | incident | 2.288 | |
-| MC-8329 | encryption | 1.790 | |
-| MC-8333 | audit_logging | 1.645 | |
-| MC-8321 | policy | 1.463 | |
-| MC-8325 | patch_management | 1.155 | |
-| MC-8338 | network_security | 1.071 | |
-| ... | (13 weitere) | 200-960 | |
+### Gap-Analyse Engine (Compliance)
+- **12 Regulierungen** automatisch klassifiziert (CRA, AI Act, NIS2, DSGVO, MiCA, PSD2, AML, etc.)
+- **IST-Zustand Assessment:** CE-Kennzeichnung, angewandte Normen, bestehende Prozesse, IACE-Projekt-Link
+- **Norm→Control Mapping:** 20 Normen → MC-Topic Coverage
+- **Prioritäts-Engine:** Severity × Deadline × Dependency
+- **5 Branchentemplates:** IoT, Exchange, Cobot, SaaS, Medical
+- **Frontend:** 2-Step Wizard (Produkt + IST-Zustand) + Dashboard mit Ampel-Status
+- **API:** 8 Endpoints unter `/sdk/v1/gap/`
+- **Persistente Projekte:** Speichern + wieder öffnen
+- **Getestet:** SmartFactory Gateway → 5 Regulierungen, 500 Gaps

-### Loesung: Regulation-Source Split
+### Tenant Document Upload API (Core)
+- `POST/GET/DELETE /api/v1/tenant/documents`
+- Tenant-isolierte Qdrant-Collections
+- Code fertig, nicht deployed (RAG Service rebuild nötig)

-Statt nur nach Embedding-Aehnlichkeit zu clustern, nach **Regulation-Quelle** aufteilen:
-
-```
-MC "encryption" (1.790 Controls)
-  → encryption_dsgvo (DSGVO Art. 32, ~200)
-  → encryption_nis2 (NIS2 Art. 21, ~150)
-  → encryption_nist (NIST SC-13, ~300)
-  → encryption_bsi (BSI, ~200)
-  → encryption_owasp (OWASP, ~100)
-  → encryption_other (~840)
-```
-
-### Script-Ansatz
-
-```python
-# Fuer jeden der 25 grossen MCs:
-# 1. Hole alle member controls mit source_citation->>'source'
-# 2. Gruppiere nach source (Regulation)
-# 3. Erstelle Sub-MCs pro Regulation
-# 4. Controls ohne source → "general" Sub-MC
-```
-
-### Qualitaetsanforderung (WICHTIG!)
-
-**Nur "sehr gut" ist akzeptabel.** Mittlere MCs (30-100 Controls) sind bereits excellent:
- MC-1082 (data_retention_policies, 52) → perfekt koharent
- MC-5477 (austausch_von_cybersicherheitsinformationen, 5) → perfekt
-
-Ziel: ALLE MCs sollen diese Qualitaet haben. Kein MC >100 Controls.
+### Master Controls Browser (Compliance)
+- **Neue Seite** `/sdk/master-controls` — reused Control Library UI
+- Sidebar-Eintrag zwischen Control Library und Provenance
+- 13.588 MCs mit allen Filtern, Paginierung, Klick-Detail
+- Verbindet sich mit Production-DB

 ---

-## SESSION 03-06.05.2026 KOMPLETT ERLEDIGT
+## DB-Tabellen (neu/geändert)

-### Block F (Hardcoded Knowledge → DB)
- F1: regulation_registry (223 Eintraege) ✅
- F2: action_types (34) + action_synonyms (368) ✅
- F3: object_synonyms (320) ✅
- F4: LLM Enrichment (+468 Synonyme via Ollama) ✅
- F5: Validation (8 Tests, Dicts als Fallback) ✅
-
-### Control Generation Pipeline
- 1.599 Rich Controls aus E-Block Chunks (~$17 Anthropic)
- 11.522 Obligations (Pass 0a, ~$4)
- 1.147 Atomic Controls (Pass 0b, ~$4.60)
- **Gesamtkosten: ~$25.60**
-
-### Production Sync
- 2.625 Controls + 11.522 Obligations auf Production synchronisiert
- Production: 294.027 Controls total
- Backups: lokal + production auf MacBook
-
-### Block G-pre (Master Controls)
- G-pre1: 144k Objects → 7.753 Gruppen (K-Means k=5000 + Sub-Cluster + Refinement)
- G-pre2: 5.329 Master Controls, 172.504+ Members
- G-pre3: Master Control API (list, stats, detail)
- **Qualitaet:** Kleine/mittlere MCs excellent, 25 grosse MCs brauchen Regulation-Source Split
-
-### Block G (Compliance Execution Layer)
- G1: Decision Trace (decision_traces Tabelle + 6 API Endpoints) ✅
- G2: Compliance Commit Ledger (compliance_commits + 5 Endpoints) ✅
- G3: Full Decision Memory (decision_events + Timeline + 4 Endpoints) ✅
- G4: Pre-Deployment Enforcement (deployment_checks + Override + 4 Endpoints) ✅
-
-### Infrastruktur
- Vault CPU-Fix committed (Marker-File + idempotente Checks)
- Pass 0a Endpoint im Core Control-Pipeline registriert
- Gitea Timezone-Fix (docker-compose.yml)
- 61 neue regulation_ids in regulation_registry
- Container-Cleanup (fewo-finance-agent, mediaanalysisd)
+| Tabelle | Repo | Rows (lokal) | Rows (Production) |
+|---------|------|-------------|-------------------|
+| compliance.master_controls | Core | 13.588 | 13.588 |
+| compliance.master_control_members | Core | 83.073 | 83.073 |
+| compliance.object_ontology | Core | 74 | 74 |
+| compliance.object_groups | Core | 16.683 | — |
+| compliance.doc_check_controls | Core | 1.874 | 1.874 |
+| compliance.gap_projects | Compliance | 1 | 0 |

 ---

-## DB-Tabellen (alle Bloecke)
+## OFFEN / NÄCHSTE SESSION

-| Tabelle | Rows | Migration |
-|---------|------|-----------|
-| compliance.regulation_registry | 223 | 002 |
-| compliance.action_types | 34 | 003 |
-| compliance.action_synonyms | 368 | 003 |
-| compliance.object_synonyms | 320 | 003 |
-| compliance.object_groups | 7.753 | 004 |
-| compliance.master_controls | 5.329 | 005 |
-| compliance.master_control_members | ~170k | 005 |
-| compliance.decision_traces | 0 (Schema ready) | 006 |
-| compliance.compliance_commits | 0 (Schema ready) | 007 |
-| compliance.decision_events | 0 (Schema ready) | 008 |
-| compliance.deployment_checks | 0 (Schema ready) | 009 |
-
---
-
-## API Endpoints (Core Control-Pipeline, Port 8098)
-
-### Bestehend
- `/v1/canonical/generate/*` — Control Generation Pipeline
- `/v1/canonical/generate/run-pass0a` — Pass 0a (NEU in dieser Session)
- `/v1/canonical/generate/submit-pass0b` — Pass 0b Batch API
-
-### Neu (diese Session)
- `/v1/master-controls` — G-pre3: Liste, Stats, Detail
- `/v1/decision-traces` — G1: CRUD + Stats
- `/v1/controls/{id}/full-trace` — G1: Volle Kette
- `/v1/compliance-commits` — G2: Commit Ledger
- `/v1/decision-events` — G3: Lifecycle Events + Timeline
- `/v1/deployment-checks` — G4: Pre-Deploy Gate + Override
-
-### API-Zugriff (WICHTIG)
-```bash
-# Nur via Docker exec (Port 8098 blockiert durch document-crawler)
-ssh macmini "/usr/local/bin/docker exec bp-core-control-pipeline curl -sf http://127.0.0.1:8098/..."
-```
+1. **Orca Deploy-Fix** — Production deployed nicht automatisch (Webhook + docker pull Problem)
+2. **Gap-Analyse v2 IST-Zustand** — Frontend Step 2 deployed, Backend deployed, aber Orca blockiert
+3. **Tenant Document Upload** deployen (RAG Service rebuild)
+4. **Compliance-Repo auf gitea pushen** — aktuell "Everything up-to-date", Orca muss manuell redeployt werden
+5. **MC-Browser erweitern** — Detail-View mit Member-Controls verbessern

 ---

 ## BACKUPS (auf MacBook)

-| Datei | Inhalt | Groesse |
-|-------|--------|---------|
-| controls_backup_20260505.csv | 1.599 neue Controls | 7.2 MB |
-| obligations_backup_20260505.csv | 11.522 Obligations | 6.2 MB |
-| production_backup_20260505.dump | Production komprimiert | 30 MB |
-| production_backup_20260505_plain.sql | Production plain | 1.3 GB |
-| local_backup_20260506.dump | Lokale DB komprimiert | ~30 MB |
-| production_backup_20260506.dump | Production komprimiert | ~30 MB |
+| Datei | Inhalt |
+|-------|--------|
+| `backup_pre_gpre3_20260510.dump` | Vor gpre3 Live-Run (171 MB) |
+| `backup_session_end_20260511.dump` | Session-Ende |
+| `production_backup_20260508.dump` | Production nach Phase 2 |
+| `gpre0_checkpoints_backup_20260508/` | 10 Corrections-JSONs |

 ---

-## GESTOPPTE CONTAINER
+## API-Kosten (Anthropic)

-```bash
-# Vault: Erst nach Fix-Deploy starten (Marker-File noetig)
-ssh macmini "/usr/local/bin/docker start bp-core-vault"
-
-# OpenSearch: Bei Bedarf
-ssh macmini "/usr/local/bin/docker start bp-lehrer-opensearch"
-
-# fewo-finance-agent: Fremder Container, nicht starten
-```
+| Phase | Modell | Kosten |
+|-------|--------|--------|
+| Phase 0: Quality Audit | Sonnet | $2.92 |
+| Phase 0b: Quality Audit v2 | Sonnet | $5.93 |
+| Phase 2: 174K Re-Klassifizierung | Haiku | ~$50 |
+| Phase 2b: Generic Token Fix | Haiku | $7.54 |
+| Phase 2c: Subtopics R1 | Haiku | $20.22 |
+| Phase 2c: Subtopics R2 | Haiku | $12.03 |
+| Phase 2d: Bad Subtopics | Haiku | ~$0.50 |
+| 5K Test-Run | Sonnet | $5.32 |
+| doc_check_controls | Haiku | ~$5 |
+| **Gesamt** | | **~$110** |

 ---

-## TESTS
+## STRATEGISCHE ENTSCHEIDUNGEN (in Memory)

-```bash
-# Pipeline (454 Tests)
-PYTHONPATH=control-pipeline python3 -m pytest control-pipeline/tests/ -v
-```
-
---
-
-## OFFENE PUNKTE FUER ANDERE SESSIONS
-
-1. **Qdrant API-Key** fuer Production (qdrant-dev.breakpilot.ai) ist ungueltig (401). Muss in Coolify erneuert werden.
-2. **DSI-Check False Positives**: Controls mischen interne Governance mit externen DSI-Anforderungen. Fix: nur Controls mit Art. 13/14 Referenz fuer DSI-Checks nutzen.
-3. **Spotlight + mediaanalysisd** auf Mac Mini deaktivieren (braucht sudo):
-   ```bash
-   sudo mdutil -a -i off
-   sudo launchctl disable system/com.apple.mediaanalysisd
-   ```
-4. **Production DB Sync** fuer neue G-Block Tabellen (decision_traces, compliance_commits, decision_events, deployment_checks) noch ausstehend — Tabellen sind leer, Schema muss auf Production deployed werden.
+1. **3 Use Cases:** Gap-Analyse (Prio 1), Vendor Risk (Prio 2), Web3/Crypto als Vertikal (Prio 3)
+2. **Keine Norm-Reproduktion:** Obligation Extraction statt ISO-Texte (juristisch sicher)
+3. **Regulatory Ingestion Engine:** BAuA/OSHA Crawler als Vorlage für automatisierte Source-Feeds
+4. **CE-Compliance Crossover:** IACE × Master Controls für Trigger-basierte Compliance-Hinweise
@@ -0,0 +1,21 @@
+import { NextRequest, NextResponse } from 'next/server'
+
+const BACKEND_URL = process.env.CONSENT_BACKEND_URL || 'https://macmini:3007/api/sdk/v1/banner'
+const TENANT_ID = process.env.CONSENT_TENANT_ID || '9282a473-5c95-4b3a-bf78-0ecc0ec71d3e'
+const SITE_ID = process.env.NEXT_PUBLIC_CONSENT_SITE_ID || 'breakpilot-marketing'
+
+export async function GET(req: NextRequest) {
+  try {
+    const siteId = req.nextUrl.searchParams.get('site_id') || SITE_ID
+    const res = await fetch(`${BACKEND_URL}/config/${siteId}`, {
+      headers: { 'X-Tenant-ID': TENANT_ID },
+    })
+    const data = await res.text()
+    return new NextResponse(data, {
+      status: res.status,
+      headers: { 'Content-Type': 'application/json' },
+    })
+  } catch {
+    return NextResponse.json({ categories: [], vendors: [] }, { status: 200 })
+  }
+}
@@ -5,7 +5,13 @@ const TENANT_ID = process.env.CONSENT_TENANT_ID || '9282a473-5c95-4b3a-bf78-0ecc

 export async function POST(req: NextRequest) {
  try {
-    const body = await req.text()
+    const data = await req.json()
+
+    // Inject client IP for backend GeoIP resolution
+    const ip = req.headers.get('x-forwarded-for')?.split(',')[0]?.trim()
+      || req.headers.get('x-real-ip')
+      || null
+    if (ip) data.ip_address = ip

    const res = await fetch(`${BACKEND_URL}/consent`, {
      method: 'POST',
@@ -13,13 +19,11 @@ export async function POST(req: NextRequest) {
        'Content-Type': 'application/json',
        'X-Tenant-ID': TENANT_ID,
      },
-      body,
-      // Accept self-signed certs on internal network
-      ...(process.env.NODE_TLS_REJECT_UNAUTHORIZED === '0' ? {} : {}),
+      body: JSON.stringify(data),
    })

-    const data = await res.text()
-    return new NextResponse(data, {
+    const resBody = await res.text()
+    return new NextResponse(resBody, {
      status: res.status,
      headers: { 'Content-Type': 'application/json' },
    })
@@ -31,7 +31,17 @@ export default function DatenschutzPage() {
            <p>
              Diese Website wird auf Servern der Hetzner Online GmbH in Deutschland gehostet.
              Es findet kein Drittlandtransfer fuer das Hosting statt.
-              Rechtsgrundlage: Art. 6 Abs. 1 lit. f DSGVO (berechtigtes Interesse an zuverlaessigem Betrieb).
+            </p>
+            <p className="mt-2">
+              <span className="text-white font-medium">Rechtsgrundlage:</span> Art. 6 Abs. 1 lit. f DSGVO (berechtigtes Interesse).
+            </p>
+            <p className="mt-1">
+              <span className="text-white font-medium">Interessenabwaegung:</span> Unser berechtigtes Interesse liegt im
+              zuverlaessigen und sicheren Betrieb der Website. Ohne Hosting-Infrastruktur koennen wir unser Angebot
+              nicht bereitstellen. Die Verarbeitung beschraenkt sich auf technisch notwendige Verbindungsdaten
+              (IP-Adresse, Zeitstempel). Entgegenstehende Interessen der Betroffenen ueberwiegen nicht, da die
+              Daten nur kurzzeitig (7 Tage) gespeichert, nicht mit anderen Datenquellen zusammengefuehrt und
+              ausschliesslich zur Sicherstellung des Betriebs und der IT-Sicherheit verwendet werden.
            </p>
          </div>

@@ -80,26 +90,29 @@ export default function DatenschutzPage() {
          <div>
            <h2 className="text-lg font-semibold text-white mb-2">5. Server-Logfiles</h2>
            <p>
-              Der Hosting-Provider erhebt technisch notwendige Logfiles (IP-Adresse, Browsertyp, Zeitstempel).
-              Rechtsgrundlage: Art. 6 Abs. 1 lit. f DSGVO (berechtigtes Interesse an der Sicherheit).
-              Logfiles werden nach 7 Tagen automatisch geloescht.
+              Der Hosting-Provider erhebt technisch notwendige Logfiles (IP-Adresse, Browsertyp, Zeitstempel,
+              aufgerufene Seite, HTTP-Statuscode).
+            </p>
+            <p className="mt-2">
+              <span className="text-white font-medium">Rechtsgrundlage:</span> Art. 6 Abs. 1 lit. f DSGVO (berechtigtes Interesse).
+            </p>
+            <p className="mt-1">
+              <span className="text-white font-medium">Interessenabwaegung:</span> Die Erhebung von Server-Logfiles ist
+              fuer die Erkennung und Abwehr von Cyberangriffen, die Fehlerbehebung und die Gewaehrleistung der
+              Systemstabilitaet unerlasslich. Die Daten werden automatisiert nach 7 Tagen geloescht und nicht
+              zur Profilbildung oder Identifizierung einzelner Nutzer verwendet. Eine Zusammenfuehrung mit anderen
+              Datenquellen findet nicht statt. Das Interesse der Betroffenen am Schutz ihrer Daten wird durch die
+              kurze Speicherdauer und die rein technische Nutzung angemessen gewahrt.
            </p>
          </div>

          <div>
-            <h2 className="text-lg font-semibold text-white mb-2">6. Externe Schriften</h2>
+            <h2 className="text-lg font-semibold text-white mb-2">6. Schriften</h2>
            <p>
-              Diese Website laedt Schriftarten von Google Fonts. Dabei wird Ihre IP-Adresse an Google LLC,
-              1600 Amphitheatre Parkway, Mountain View, CA 94043, USA uebermittelt.
-              Rechtsgrundlage: Art. 6 Abs. 1 lit. f DSGVO.
-              Google ist unter dem EU-US Data Privacy Framework (DPF) zertifiziert (Angemessenheitsbeschluss
-              der EU-Kommission vom 10. Juli 2023).
-            </p>
-            <p className="mt-1">
-              Weitere Informationen:{' '}
-              <a href="https://policies.google.com/privacy" className="text-accent-electric hover:underline" target="_blank" rel="noopener noreferrer">
-                policies.google.com/privacy
-              </a>
+              Diese Website verwendet die Schriftarten Inter, Plus Jakarta Sans und JetBrains Mono.
+              Die Schriften werden lokal auf unserem Server gehostet — es findet kein Abruf von
+              externen Servern (z.B. Google Fonts) statt. Es werden keine personenbezogenen Daten
+              an Dritte uebermittelt.
            </p>
          </div>

@@ -116,8 +129,8 @@ export default function DatenschutzPage() {
            <h2 className="text-lg font-semibold text-white mb-2">8. Empfaenger und Auftragsverarbeiter</h2>
            <ul className="list-disc list-inside space-y-1">
              <li>Hetzner Online GmbH, Industriestr. 25, 91710 Gunzenhausen — Hosting (AVV nach Art. 28 DSGVO)</li>
-              <li>Google LLC — Schriftarten (Google Fonts CDN, EU-US DPF)</li>
            </ul>
+            <p className="mt-1">Schriftarten werden lokal gehostet — kein Drittanbieter-Transfer.</p>
          </div>

          <div>
@@ -1,6 +1,28 @@
-@import url('https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600;700;800;900&display=swap');
-@import url('https://fonts.googleapis.com/css2?family=Plus+Jakarta+Sans:wght@400;500;600;700;800&display=swap');
-@import url('https://fonts.googleapis.com/css2?family=JetBrains+Mono:wght@400;500;600&display=swap');
+/* Self-hosted fonts — kein Drittlandtransfer zu Google */
+@font-face {
+  font-family: 'Inter';
+  font-style: normal;
+  font-weight: 300 900;
+  font-display: swap;
+  src: url('/fonts/Inter-Latin.woff2') format('woff2');
+  unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+0304, U+0308, U+0329, U+2000-206F, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;
+}
+@font-face {
+  font-family: 'Plus Jakarta Sans';
+  font-style: normal;
+  font-weight: 400 800;
+  font-display: swap;
+  src: url('/fonts/PlusJakartaSans-Latin.woff2') format('woff2');
+  unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+0304, U+0308, U+0329, U+2000-206F, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;
+}
+@font-face {
+  font-family: 'JetBrains Mono';
+  font-style: normal;
+  font-weight: 400 600;
+  font-display: swap;
+  src: url('/fonts/JetBrainsMono-Latin.woff2') format('woff2');
+  unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+0304, U+0308, U+0329, U+2000-206F, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;
+}

@tailwind base;
@tailwind components;
@@ -1,6 +1,7 @@
 import type { Metadata } from 'next'
 import { AppProvider } from '@/lib/context'
 import ConsentBanner from '@/components/layout/ConsentBanner'
+import ScriptManager from '@/components/layout/ScriptManager'
 import './globals.css'

 export const metadata: Metadata = {
@@ -27,6 +28,7 @@ export default function RootLayout({
        <AppProvider>
          {children}
          <ConsentBanner />
+          <ScriptManager />
        </AppProvider>
      </body>
    </html>
@@ -51,6 +51,18 @@ const texts = {
  },
 }

+function getSessionId(): string {
+  if (typeof window === 'undefined') return ''
+  try {
+    let sid = sessionStorage.getItem('bp_session_id')
+    if (!sid) {
+      sid = crypto.randomUUID()
+      sessionStorage.setItem('bp_session_id', sid)
+    }
+    return sid
+  } catch { return '' }
+}
+
 function getFingerprint(): string {
  const nav = typeof navigator !== 'undefined' ? navigator : null
  const raw = [nav?.language, nav?.platform, screen?.width, screen?.height, new Date().getTimezoneOffset()].join('|')
@@ -86,9 +98,40 @@ function detectDevice(): { device_type: string; browser: string; os: string } {

 type ConsentMethod = 'accept_all' | 'reject_all' | 'custom_selection'

-async function sendConsent(consent: ConsentState, method: ConsentMethod) {
+interface ScriptEntry { src: string; category: string }
+interface CookieEntry { name: string; domain: string; expiry_days: number; category: string }
+
+function detectScripts(): { blocked: ScriptEntry[]; released: ScriptEntry[] } {
+  const scripts = Array.from(document.querySelectorAll('script[src]'))
+  const released: ScriptEntry[] = []
+  const blocked: ScriptEntry[] = []
+  for (const el of scripts) {
+    const src = el.getAttribute('src') || ''
+    if (/google.*tag|gtag|analytics/i.test(src)) released.push({ src, category: 'analytics' })
+    else if (/facebook|fbevents|linkedin|tiktok/i.test(src)) released.push({ src, category: 'marketing' })
+  }
+  return { blocked, released }
+}
+
+function detectCookies(): CookieEntry[] {
+  const cookies: CookieEntry[] = []
+  for (const c of document.cookie.split(';')) {
+    const name = c.trim().split('=')[0]
+    if (!name) continue
+    let category = 'functional'
+    if (/^_ga|^_gid|^_gat/i.test(name)) category = 'analytics'
+    else if (/^_fb|^_gcl|^_li/i.test(name)) category = 'marketing'
+    else if (/^bp_consent|^session|^csrf/i.test(name)) category = 'essential'
+    cookies.push({ name, domain: window.location.hostname, expiry_days: 0, category })
+  }
+  return cookies
+}
+
+async function sendConsent(consent: ConsentState, method: ConsentMethod, vendorConsents?: Record<string, boolean>) {
  try {
    const { device_type, browser, os } = detectDevice()
+    const { blocked, released } = detectScripts()
+    const cookies_set = detectCookies()
    await fetch('/api/consent', {
      method: 'POST',
      headers: { 'Content-Type': 'application/json' },
@@ -100,7 +143,8 @@ async function sendConsent(consent: ConsentState, method: ConsentMethod) {
          ...(consent.functional ? ['functional'] : []),
          ...(consent.analytics ? ['analytics'] : []),
        ],
-        vendors: [],
+        vendors: Object.keys(vendorConsents || {}).filter(k => vendorConsents?.[k]),
+        vendor_consents: vendorConsents || {},
        user_agent: navigator.userAgent,
        consent_method: method,
        page_url: window.location.href,
@@ -110,6 +154,11 @@ async function sendConsent(consent: ConsentState, method: ConsentMethod) {
        os,
        screen_resolution: `${screen.width}x${screen.height}`,
        consent_scope: 'domain',
+        session_id: getSessionId(),
+        timezone: Intl.DateTimeFormat().resolvedOptions().timeZone,
+        scripts_blocked: blocked,
+        scripts_released: released,
+        cookies_set,
      }),
    })
  } catch {
@@ -117,26 +166,49 @@ async function sendConsent(consent: ConsentState, method: ConsentMethod) {
  }
 }

+interface VendorConfig {
+  vendor_name: string
+  category_key: string
+  description_de?: string
+  description_en?: string
+  cookie_names?: string[]
+  retention_days?: number
+}
+
 export default function ConsentBanner() {
  const { lang } = useApp()
  const t = texts[lang]
  const [visible, setVisible] = useState(false)
  const [showDetails, setShowDetails] = useState(false)
  const [consent, setConsent] = useState<ConsentState>(defaultConsent)
+  const [vendors, setVendors] = useState<VendorConfig[]>([])
+  const [vendorConsents, setVendorConsents] = useState<Record<string, boolean>>({})

  useEffect(() => {
    const saved = getSavedConsent()
    if (!saved) {
      setVisible(true)
    }
+    // Load vendor config from backend
+    fetch('/api/consent/config')
+      .then(r => r.json())
+      .then(data => {
+        const v = data?.vendors || []
+        setVendors(v)
+        // Default all vendors to true
+        const defaults: Record<string, boolean> = {}
+        for (const vendor of v) defaults[vendor.vendor_name] = true
+        setVendorConsents(defaults)
+      })
+      .catch(() => {})
  }, [])

  const save = useCallback((state: ConsentState, method: ConsentMethod) => {
    localStorage.setItem(COOKIE_NAME, JSON.stringify(state))
-    sendConsent(state, method)
+    sendConsent(state, method, vendorConsents)
    setVisible(false)
    window.dispatchEvent(new CustomEvent('consent-change', { detail: state }))
-  }, [])
+  }, [vendorConsents])

  const acceptAll = () => save({ essential: true, functional: true, analytics: true }, 'accept_all')
  const rejectAll = () => save({ essential: true, functional: false, analytics: false }, 'reject_all')
@@ -181,24 +253,49 @@ export default function ConsentBanner() {
                className="overflow-hidden mb-4"
              >
                <div className="space-y-2 pt-2 border-t border-white/[0.06]">
-                  {categories.map(([key, cat]) => (
-                    <label
-                      key={key}
-                      className="flex items-center justify-between p-3 rounded-xl bg-white/[0.03] border border-white/[0.06] cursor-pointer hover:bg-white/[0.05] transition-colors"
-                    >
-                      <div className="flex-1 mr-4">
-                        <span className="text-xs font-semibold text-white">{cat.name}</span>
-                        <p className="text-xs text-white/40 mt-0.5">{cat.description}</p>
+                  {categories.map(([key, cat]) => {
+                    const catVendors = vendors.filter(v => v.category_key === key)
+                    return (
+                      <div key={key} className="rounded-xl bg-white/[0.03] border border-white/[0.06]">
+                        <label className="flex items-center justify-between p-3 cursor-pointer hover:bg-white/[0.05] transition-colors">
+                          <div className="flex-1 mr-4">
+                            <span className="text-xs font-semibold text-white">{cat.name}</span>
+                            <p className="text-xs text-white/40 mt-0.5">{cat.description}</p>
+                          </div>
+                          <input
+                            type="checkbox"
+                            checked={cat.required || consent[key as keyof ConsentState]}
+                            disabled={cat.required}
+                            onChange={(e) => {
+                              setConsent(prev => ({ ...prev, [key]: e.target.checked }))
+                              for (const v of catVendors) {
+                                setVendorConsents(prev => ({ ...prev, [v.vendor_name]: e.target.checked }))
+                              }
+                            }}
+                            className="w-4 h-4 rounded accent-accent-electric"
+                          />
+                        </label>
+                        {catVendors.length > 0 && consent[key as keyof ConsentState] && (
+                          <div className="px-3 pb-3 space-y-1">
+                            {catVendors.map(v => (
+                              <label key={v.vendor_name} className="flex items-center justify-between pl-4 py-1 text-xs cursor-pointer">
+                                <div className="flex-1 mr-2">
+                                  <span className="text-white/60">{v.vendor_name}</span>
+                                  {v.retention_days && <span className="text-white/30 ml-1">({v.retention_days}d)</span>}
+                                </div>
+                                <input
+                                  type="checkbox"
+                                  checked={vendorConsents[v.vendor_name] ?? true}
+                                  onChange={(e) => setVendorConsents(prev => ({ ...prev, [v.vendor_name]: e.target.checked }))}
+                                  className="w-3 h-3 rounded accent-accent-electric"
+                                />
+                              </label>
+                            ))}
+                          </div>
+                        )}
                      </div>
-                      <input
-                        type="checkbox"
-                        checked={cat.required || consent[key as keyof ConsentState]}
-                        disabled={cat.required}
-                        onChange={(e) => setConsent(prev => ({ ...prev, [key]: e.target.checked }))}
-                        className="w-4 h-4 rounded accent-accent-electric"
-                      />
-                    </label>
-                  ))}
+                    )
+                  })}
                </div>
              </motion.div>
            )}
@@ -0,0 +1,115 @@
+'use client'
+
+import { useEffect, useRef, useCallback } from 'react'
+
+/**
+ * ScriptManager — active consent-aware script blocking + injection.
+ *
+ * Two mechanisms:
+ * 1. INJECTION: Scripts in CONSENT_SCRIPTS are only injected AFTER consent.
+ * 2. BLOCKING: Existing <script data-consent="category" type="text/plain">
+ *    elements in the page are activated after consent by changing type to
+ *    "text/javascript". This is the standard CMP blocking pattern.
+ *
+ * Usage for inline scripts in pages:
+ *   <script type="text/plain" data-consent="analytics">
+ *     // This won't execute until analytics consent is given
+ *     gtag('config', 'G-XXXXXX');
+ *   </script>
+ *
+ * Usage for adding new third-party scripts:
+ *   Add to CONSENT_SCRIPTS below. They'll be injected only after consent.
+ */
+
+interface ConsentScript {
+  src: string
+  async?: boolean
+  id?: string
+}
+
+const CONSENT_SCRIPTS: Record<string, ConsentScript[]> = {
+  analytics: [
+    // { src: 'https://www.googletagmanager.com/gtag/js?id=G-XXXXXX', async: true, id: 'gtag' },
+    // { src: 'https://plausible.io/js/script.js', async: true, id: 'plausible' },
+  ],
+  marketing: [
+    // { src: 'https://connect.facebook.net/en_US/fbevents.js', async: true, id: 'fb-pixel' },
+    // { src: 'https://snap.licdn.com/li.lms-analytics/insight.min.js', async: true, id: 'li-insight' },
+  ],
+  functional: [
+    // { src: 'https://widget.example.com/chat.js', async: true, id: 'chat-widget' },
+  ],
+}
+
+interface ConsentState {
+  essential: boolean
+  functional: boolean
+  analytics: boolean
+}
+
+function getStoredConsent(): ConsentState | null {
+  if (typeof window === 'undefined') return null
+  try {
+    const raw = localStorage.getItem('bp_consent')
+    return raw ? JSON.parse(raw) : null
+  } catch { return null }
+}
+
+export default function ScriptManager() {
+  const injected = useRef(new Set<string>())
+
+  const applyConsent = useCallback((consent: ConsentState) => {
+    const accepted = new Set<string>()
+    accepted.add('essential') // always allowed
+    if (consent.functional) accepted.add('functional')
+    if (consent.analytics) accepted.add('analytics')
+
+    // 1. INJECT: Add scripts from CONSENT_SCRIPTS for accepted categories
+    for (const cat of accepted) {
+      for (const script of CONSENT_SCRIPTS[cat] ?? []) {
+        if (injected.current.has(script.src)) continue
+        const el = document.createElement('script')
+        el.src = script.src
+        if (script.async) el.async = true
+        if (script.id) el.id = script.id
+        el.dataset.consent = cat
+        document.head.appendChild(el)
+        injected.current.add(script.src)
+      }
+    }
+
+    // 2. ACTIVATE: Unblock <script type="text/plain" data-consent="...">
+    const blocked = document.querySelectorAll('script[type="text/plain"][data-consent]')
+    for (const el of blocked) {
+      const cat = el.getAttribute('data-consent') || ''
+      if (accepted.has(cat)) {
+        const clone = document.createElement('script')
+        // Copy attributes
+        for (const attr of el.attributes) {
+          if (attr.name === 'type') continue // skip type="text/plain"
+          clone.setAttribute(attr.name, attr.value)
+        }
+        clone.type = 'text/javascript'
+        clone.textContent = el.textContent
+        el.parentNode?.replaceChild(clone, el)
+      }
+    }
+  }, [])
+
+  useEffect(() => {
+    // On mount: apply saved consent (return visitors)
+    const saved = getStoredConsent()
+    if (saved) applyConsent(saved)
+
+    // Listen for new consent decisions
+    function onConsentChange(e: Event) {
+      const detail = (e as CustomEvent<ConsentState>).detail
+      if (detail) applyConsent(detail)
+    }
+
+    window.addEventListener('consent-change', onConsentChange)
+    return () => window.removeEventListener('consent-change', onConsentChange)
+  }, [applyConsent])
+
+  return null
+}