Files
breakpilot-compliance/backend-compliance/compliance/services/banner_cookie_cross_check.py
T
Benjamin Admin c867478791
CI / loc-budget (push) Failing after 16s
Build + Deploy / build-admin-compliance (push) Successful in 14s
Build + Deploy / build-backend-compliance (push) Successful in 16s
Build + Deploy / build-ai-sdk (push) Successful in 20s
Build + Deploy / build-developer-portal (push) Successful in 12s
Build + Deploy / build-tts (push) Successful in 15s
Build + Deploy / build-document-crawler (push) Successful in 13s
Build + Deploy / build-dsms-gateway (push) Successful in 13s
Build + Deploy / build-dsms-node (push) Successful in 12s
CI / branch-name (push) Has been skipped
CI / guardrail-integrity (push) Has been skipped
CI / test-python-document-crawler (push) Successful in 26s
CI / secret-scan (push) Has been skipped
CI / go-lint (push) Has been skipped
CI / python-lint (push) Has been skipped
CI / nodejs-lint (push) Has been skipped
CI / nodejs-build (push) Successful in 2m49s
CI / dep-audit (push) Has been skipped
CI / sbom-scan (push) Has been skipped
CI / test-go (push) Successful in 45s
CI / test-python-backend (push) Successful in 38s
CI / test-python-dsms-gateway (push) Successful in 23s
CI / validate-canonical-controls (push) Successful in 15s
Build + Deploy / trigger-orca (push) Successful in 2m23s
feat(tcf-vendors): GVL cache + vendor extraction + VVT mapping
Phase 1-2 of the closed quality loop:
- GVL cache (consent-tester/services/gvl_cache.py): downloads and caches
  IAB Global Vendor List with 24h TTL, resolves vendor IDs to names,
  purposes, policy URLs, retention, country
- Vendor extraction (consent_interceptor.py): extract_tcf_vendors()
  reads __tcfapi after accept phase, resolves via GVL
- Scan response: tcf_vendors field added to /scan endpoint
- VVT mapper (vendor_vvt_mapper.py): maps TCF vendors to VVT format
  with purpose labels, Rechtsgrundlage, Drittland detection
- Vendor cross-check (banner_cookie_cross_check.py): checks all TCF
  vendors against DSI text — missing vendors, undocumented transfers
- Compliance check integrates Step 3d: TCF vendors vs DSI

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-12 18:18:50 +02:00

226 lines
8.7 KiB
Python

"""
Banner ↔ Cookie-Richtlinie Cross-Check.
Compares what the cookie banner actually does (Playwright scan)
with what the cookie policy text promises. Finds mismatches:
- Services used but not documented
- Tracking before consent despite claiming consent-only
- False "no tracking" claims contradicted by actual cookies
"""
import logging
import re
logger = logging.getLogger(__name__)
def cross_check_banner_vs_cookie(
banner_result: dict,
cookie_text: str,
) -> list[dict]:
"""Cross-check: Does the banner behavior match the cookie policy text?
Returns list of CheckItem-compatible dicts for mismatches.
"""
findings: list[dict] = []
cookie_lower = cookie_text.lower()
phases = banner_result.get("phases", {})
before = phases.get("before_consent", {})
after_accept = phases.get("after_accept", {})
# 1. Services found by banner but not mentioned in cookie policy
tracking_before = before.get("tracking_services", [])
all_tracking = set(tracking_before)
for s in after_accept.get("tracking_services", []):
if isinstance(s, str):
all_tracking.add(s)
for service in all_tracking:
service_lower = service.lower()
mentioned = any(kw in cookie_lower for kw in [
service_lower,
service_lower.replace(" ", ""),
service_lower.split()[0] if " " in service_lower else service_lower,
])
if not mentioned:
findings.append({
"id": f"cross-{service_lower.replace(' ', '_')[:20]}",
"label": f"Dienst '{service}' fehlt in Cookie-Richtlinie",
"passed": False,
"severity": "HIGH",
"level": 2,
"parent": None,
"skipped": False,
"matched_text": "",
"hint": (
f"Der Banner-Scan hat '{service}' auf der Website erkannt, "
f"aber die Cookie-Richtlinie erwaehnt diesen Dienst nicht. "
f"Gemaess Art. 13(1)(e) DSGVO muessen alle Empfaenger/"
f"Dienste in der DSI benannt werden."
),
"source": "cross_check",
})
# 2. Tracking before consent but cookie policy says "nur mit Einwilligung"
if tracking_before:
consent_words = re.findall(
r"(?:einwilligung|consent|zustimmung).*?"
r"(?:erforderlich|notwendig|vorher|vorab)",
cookie_lower,
)
if consent_words:
findings.append({
"id": "cross-tracking-before-consent",
"label": "Tracking vor Consent trotz Einwilligungspflicht",
"passed": False,
"severity": "CRITICAL",
"level": 2,
"parent": None,
"skipped": False,
"matched_text": ", ".join(tracking_before[:3]),
"hint": (
f"Die Cookie-Richtlinie spricht von Einwilligungspflicht, "
f"aber der Banner-Scan hat {len(tracking_before)} Tracking-"
f"Dienst(e) VOR Consent erkannt: {', '.join(tracking_before[:3])}. "
f"Dies ist ein Verstoss gegen §25 Abs. 1 TDDDG."
),
"source": "cross_check",
})
# 3. Banner sets tracking cookies but policy claims "no tracking"
accept_cookies = after_accept.get("cookies", [])
tracking_prefixes = ("_ga", "_gid", "_fbp", "_fbc", "IDE", "_gcl",
"fr", "_pin", "_tt_", "li_sugr", "_hj")
tracking_cookies = [c for c in accept_cookies
if any(c.startswith(p) for p in tracking_prefixes)]
if tracking_cookies:
no_tracking_claim = any(kw in cookie_lower for kw in [
"keine tracking", "kein tracking", "keine marketing",
"keine werbe", "no tracking",
])
if no_tracking_claim:
findings.append({
"id": "cross-false-no-tracking",
"label": "Cookie-Richtlinie behauptet 'kein Tracking'"
"Banner widerspricht",
"passed": False,
"severity": "CRITICAL",
"level": 2,
"parent": None,
"skipped": False,
"matched_text": ", ".join(tracking_cookies[:5]),
"hint": (
f"Die Cookie-Richtlinie behauptet kein Tracking, aber "
f"nach Akzeptieren wurden Tracking-Cookies gesetzt: "
f"{', '.join(tracking_cookies[:5])}. "
f"Transparenzverstoss nach Art. 5(1)(a) DSGVO."
),
"source": "cross_check",
})
# 4. Banner detected but no cookie policy provided
banner_detected = banner_result.get("banner_detected", False)
if banner_detected and len(cookie_lower.strip()) < 100:
findings.append({
"id": "cross-banner-no-policy",
"label": "Cookie-Banner vorhanden aber keine Cookie-Richtlinie",
"passed": False,
"severity": "HIGH",
"level": 2,
"parent": None,
"skipped": False,
"matched_text": "",
"hint": (
"Die Website zeigt ein Cookie-Banner, aber es wurde keine "
"Cookie-Richtlinie eingereicht. Eine separate Cookie-Richtlinie "
"oder ein Cookie-Abschnitt in der DSI ist nach §25 TDDDG "
"erforderlich wenn nicht-essentielle Cookies verwendet werden."
),
"source": "cross_check",
})
logger.info("Cross-check: %d findings (%d services, %d tracking before)",
len(findings), len(all_tracking), len(tracking_before))
return findings
def cross_check_vendors_vs_dsi(
vendors: list[dict],
dsi_text: str,
) -> list[dict]:
"""Cross-check: Are all TCF vendors documented in the DSI?
Checks per vendor:
1. Is the vendor mentioned by name?
2. Is third-country transfer documented (if non-EU)?
3. Is storage duration mentioned?
Returns list of CheckItem-compatible dicts.
"""
findings: list[dict] = []
dsi_lower = dsi_text.lower()
for v in vendors:
name = v.get("name", "")
name_lower = name.lower()
if not name_lower:
continue
# Check if vendor is mentioned in DSI
mentioned = any(kw in dsi_lower for kw in [
name_lower,
name_lower.replace(" ", ""),
name_lower.split()[0] if " " in name_lower else name_lower,
])
if not mentioned:
findings.append({
"id": f"vendor-{v.get('vendor_id', name_lower[:20])}",
"label": f"Verarbeiter '{name}' fehlt in DSI",
"passed": False,
"severity": "HIGH",
"level": 2,
"parent": None,
"skipped": False,
"matched_text": "",
"hint": (
f"Der Cookie-Banner listet '{name}' als Verarbeiter "
f"({v.get('zweck_kurz', 'unbekannt')}), aber die DSI "
f"erwaehnt diesen Dienst nicht. Art. 13(1)(e) DSGVO "
f"verlangt die Benennung aller Empfaenger."
),
"source": "vendor_cross_check",
})
# Check third-country transfer documentation
if v.get("drittland") and mentioned:
country = v.get("land", "Drittland")
transfer_mentioned = any(kw in dsi_lower for kw in [
name_lower + ".*" + "usa",
name_lower + ".*" + "drittland",
"scc", "standardvertragsklausel", "data privacy framework",
"angemessenheitsbeschluss",
])
if not transfer_mentioned:
findings.append({
"id": f"vendor-transfer-{v.get('vendor_id', '')}",
"label": f"Drittlandtransfer fuer '{name}' nicht dokumentiert",
"passed": False,
"severity": "MEDIUM",
"level": 2,
"parent": None,
"skipped": False,
"matched_text": "",
"hint": (
f"'{name}' verarbeitet Daten in {country} (ausserhalb EWR). "
f"Die DSI muss den Transfermechanismus benennen: "
f"SCC (Art. 46(2)(c)) oder DPF (Angemessenheitsbeschluss)."
),
"source": "vendor_cross_check",
})
logger.info("Vendor cross-check: %d findings for %d vendors",
len(findings), len(vendors))
return findings