feat: Document Templates V2 — DSFA, TOM, VVT, AVV, Verpflichtung, Art.13/14

Erweiterte Compliance-Vorlagen fuer den Document Generator:
- DSFA V2: Schwellwertanalyse (9 WP248-Kriterien), SDM-basierte TOM,
  strukturierte Risikobewertung, KI-Modul (AI Act), Art.36-Pruefung
- TOM V2: 7 SDM-Gewaehrleistungsziele, Sektor-Erweiterungen,
  NIS2/ISO27001/AI Act Varianten
- VVT V2: 6 Branchen-Muster (IT/SaaS, Gesundheit, Handel, Handwerk,
  Bildung, Beratung) + allgemeine Art.30-Vorlage
- AVV V2: Vollstaendiger Art.28-Vertrag mit TOM-Anlage
- Verpflichtungserklaerung: Mitarbeiter-Vertraulichkeit
- Art.13/14 Informationspflichten-Muster

Enthalt SQL-Migrations (compliance_legal_templates), Python-Generatoren
und Qdrant-Cleanup-Skript. Feature-Branch fuer spaetere Integration
in breakpilot-compliance.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

document-templates/scripts/cleanup_temp_vorlagen.py

@@ -0,0 +1,137 @@
+#!/usr/bin/env python3
+"""Cleanup script: Delete temporary DPA template documents from Qdrant.
+
+Removes all points with payload field `temp_vorlagen=true` from
+the bp_compliance_datenschutz collection.
+
+Usage:
+    python cleanup_temp_vorlagen.py --dry-run     # Preview only
+    python cleanup_temp_vorlagen.py               # Execute deletion
+    python cleanup_temp_vorlagen.py --qdrant-url http://localhost:6333
+"""
+
+import argparse
+import json
+import sys
+from typing import Optional
+from urllib.request import Request, urlopen
+from urllib.error import URLError
+
+
+def qdrant_request(base_url: str, method: str, path: str, body: Optional[dict] = None) -> dict:
+    url = f"{base_url}{path}"
+    data = json.dumps(body).encode() if body else None
+    headers = {"Content-Type": "application/json"} if data else {}
+    req = Request(url, data=data, headers=headers, method=method)
+    with urlopen(req, timeout=30) as resp:
+        return json.loads(resp.read())
+
+
+def count_temp_vorlagen(base_url: str, collection: str) -> int:
+    """Count points with temp_vorlagen=true."""
+    body = {
+        "filter": {
+            "must": [
+                {"key": "temp_vorlagen", "match": {"value": True}}
+            ]
+        },
+        "limit": 0,
+        "exact": True,
+    }
+    result = qdrant_request(base_url, "POST", f"/collections/{collection}/points/count", body)
+    return result.get("result", {}).get("count", 0)
+
+
+def list_temp_regulation_ids(base_url: str, collection: str) -> list[str]:
+    """Get distinct regulation_ids of temp documents."""
+    body = {
+        "filter": {
+            "must": [
+                {"key": "temp_vorlagen", "match": {"value": True}}
+            ]
+        },
+        "limit": 500,
+        "with_payload": ["regulation_id", "title", "source"],
+    }
+    result = qdrant_request(base_url, "POST", f"/collections/{collection}/points/scroll", body)
+    points = result.get("result", {}).get("points", [])
+
+    seen = {}
+    for p in points:
+        payload = p.get("payload", {})
+        rid = payload.get("regulation_id", "unknown")
+        if rid not in seen:
+            seen[rid] = {
+                "regulation_id": rid,
+                "title": payload.get("title", ""),
+                "source": payload.get("source", ""),
+            }
+    return list(seen.values())
+
+
+def delete_temp_vorlagen(base_url: str, collection: str) -> int:
+    """Delete all points with temp_vorlagen=true."""
+    body = {
+        "filter": {
+            "must": [
+                {"key": "temp_vorlagen", "match": {"value": True}}
+            ]
+        }
+    }
+    result = qdrant_request(base_url, "POST", f"/collections/{collection}/points/delete", body)
+    status = result.get("status", "unknown")
+    return status
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Delete temp DPA templates from Qdrant")
+    parser.add_argument("--qdrant-url", default="http://localhost:6333",
+                        help="Qdrant URL (default: http://localhost:6333)")
+    parser.add_argument("--collection", default="bp_compliance_datenschutz",
+                        help="Qdrant collection name")
+    parser.add_argument("--dry-run", action="store_true",
+                        help="Only count and list, do not delete")
+    args = parser.parse_args()
+
+    print(f"Qdrant URL: {args.qdrant_url}")
+    print(f"Collection: {args.collection}")
+    print()
+
+    try:
+        count = count_temp_vorlagen(args.qdrant_url, args.collection)
+    except URLError as e:
+        print(f"ERROR: Cannot connect to Qdrant at {args.qdrant_url}: {e}")
+        sys.exit(1)
+
+    print(f"Gefundene Punkte mit temp_vorlagen=true: {count}")
+
+    if count == 0:
+        print("Nichts zu loeschen.")
+        return
+
+    docs = list_temp_regulation_ids(args.qdrant_url, args.collection)
+    print(f"\nBetroffene Dokumente ({len(docs)}):")
+    for doc in sorted(docs, key=lambda d: d["regulation_id"]):
+        source = f" [{doc['source']}]" if doc.get("source") else ""
+        title = f" — {doc['title']}" if doc.get("title") else ""
+        print(f"  - {doc['regulation_id']}{title}{source}")
+
+    if args.dry_run:
+        print(f"\n[DRY-RUN] Wuerde {count} Punkte loeschen. Keine Aenderung durchgefuehrt.")
+        return
+
+    print(f"\nLoesche {count} Punkte ...")
+    status = delete_temp_vorlagen(args.qdrant_url, args.collection)
+    print(f"Status: {status}")
+
+    remaining = count_temp_vorlagen(args.qdrant_url, args.collection)
+    print(f"Verbleibende temp_vorlagen Punkte: {remaining}")
+
+    if remaining == 0:
+        print("Cleanup erfolgreich abgeschlossen.")
+    else:
+        print(f"WARNUNG: {remaining} Punkte konnten nicht geloescht werden.")
+
+
+if __name__ == "__main__":
+    main()