Commit Graph

645 Commits

Author SHA1 Message Date
Benjamin Admin 1b37b2aeea refactor: Cookie banner — categories always visible (CNIL/DSK compliant)
- All 4 categories with toggles visible on first layer (no "Einstellungen" step)
- Removed showSettings state — single-view banner
- EWR toggle + info button in header, always visible
- Two equal-weight buttons: "Alle akzeptieren" + "Auswahl speichern"
- "Nur notwendige" as text link below (not hidden, but less prominent)
- Vendor tables expandable per category via chevron
- DSK OH Telemedien 2022 + CNIL 2020 compliant layout

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 22:36:27 +02:00
Benjamin Admin 4a688098e8 fix: Move EWR toggle to banner header with info button
- EWR toggle now visible on initial banner view (top-right, always visible)
- Info button (i) with tooltip explaining EWR-only mode
- Blocked vendors count badge below toggle
- Blocked vendor pills shown below header text
- Removed duplicate EWR section from settings view

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 22:18:45 +02:00
Benjamin Admin a2492f0b7e feat: "Nur EU/EWR" toggle in Cookie Banner — blocks non-EWR vendors
Game-changing CMP feature: Users accept a category (e.g. Marketing) but
can restrict data processing to EU/EWR-only vendors. Non-EWR vendors are
blocked even when the category is accepted.

- Toggle "Nur EU/EWR-Anbieter" with globe icon in blue gradient bar
- Blocked vendors shown as red pills with strikethrough icon
- Per-vendor status icons: green checkmark (active), red slash (blocked),
  gray dash (category disabled)
- Country column: green circle+check for EWR, amber warning for non-EWR
- EWR = EU27 + IS/LI/NO + CH (Angemessenheitsbeschluss)
- Vendor data extracted to cookie-banner-vendors.ts (under 500 LOC)
- Consent state includes ewrOnly flag + blockedVendors list

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 21:26:45 +02:00
Benjamin Admin fe6764df9a fix: ensure JSONB array fields are always arrays in control API
Backend: _ensure_list() converts null/string/malformed JSONB to []
for requirements, test_procedure, evidence, open_anchors, tags.

Frontend: defensive Array.isArray() check on ControlDetail.tsx.

Fixes: TypeError: A.requirements.map is not a function

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 21:18:10 +02:00
Benjamin Admin db697924ed feat: Cookie banner vendors per category + {{COOKIE_TABLE}} generator
- CookieBannerOverlay: shows vendors per category with expandable tables
  (Verarbeiter, Cookies, Dauer, Land) for full transparency
- Demo vendors: 4 necessary, 3 statistics, 3 marketing, 3 functional
- cookie_table_generator.py: renders {{COOKIE_TABLE}} Markdown tables
  from vendor configs (DB) or service registry (fallback)
- SERVICE_COOKIES: 16 known vendor-to-cookie mappings with provider + country

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 20:06:44 +02:00
Benjamin Admin f9a1fe21dc feat: Live cookie banner overlay in SDK — auto-open + FAB reopen button
- CookieBannerOverlay: opens automatically on first visit (localStorage check)
- CookieBannerFAB: shield icon button at right-[10rem] to reopen settings
- 3 consent modes: accept all, reject all (nur notwendige), custom settings
- 4 categories: Notwendig (locked on), Statistik, Marketing, Funktional
- Category toggles with descriptions in settings view
- Datenschutzerklaerung + Impressum links in banner
- Consent persisted to localStorage, custom event fired on change
- Comprehensive Playwright E2E tests (16 tests):
  - First visit auto-open, button visibility, category toggles
  - Accept all / reject all / custom settings persistence
  - FAB reopen behavior, disabled toggle for necessary category

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 19:46:22 +02:00
Benjamin Admin 17c67b4f25 feat: Cookie-Banner ↔ Backend Integration (DSR, Retention, Consent Proof)
Phase 1: Vendor sync from service registry (82+ services → banner vendors)
Phase 2: Category-based retention (marketing=90d, statistics=790d, not hardcoded 365d)
Phase 3: DSR ↔ Banner email linking (link-email, by-email, Art.17 erasure, Art.15/20 export)
Phase 4: Consent sync (Banner → Einwilligungen bridge)
Phase 6: Consent proof (SHA256 config hash + config_version in audit log, Art. 7(1) DSGVO)

New files:
- banner_dsr_service.py — email linking + DSR integration
- vendor_banner_sync.py — service registry → vendor configs
- migration 106 — linked_email, banner_config_hash, consent_version columns

Tests: 20+ new backend tests + 2 Playwright E2E test suites (API + UI)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 19:41:22 +02:00
Benjamin Admin cb2d503e84 feat: Google Consent Mode v2 + Developer Portal cookie banner docs
Phase A: Google Consent Mode v2 in cookie-banner-embed.ts
- gtag('consent', 'default', {...denied}) before banner loads
- gtag('consent', 'update', {...}) after user decision
- Automatic mapping: statistics→analytics_storage, marketing→ad_storage

Phase B: 5 Developer Portal pages at /sdk/consent/cookie-banner/
- Overview page with 4 cards
- Integration Guide: 3-step setup, script-tag, categories
- Google Consent Mode: automatic integration, parameter mapping
- Script Blocking: type=text/plain pattern, GA/FB/Hotjar examples
- Compliance Checklist: 12 points, 9 automatic

Sidebar navigation extended with Cookie-Banner section.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 17:13:34 +02:00
Benjamin Admin dccd9d09e5 feat: cookie banner compliance hardening — 5 legal requirements
1. Impressum link mandatory in banner (§5 TMG)
2. Pre-ticked prevention: only "required" categories pre-enabled (Planet49)
3. Cookie-Settings reopen link (§7(3) DSGVO — revocation as easy as consent)
4. Script-Blocking: data-cookie-category + type="text/plain" pattern
   Scripts only execute AFTER user consents to that category
5. Buttons already equal size (flex:1) — verified correct

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 15:50:54 +02:00
Benjamin Admin ca21feedc8 feat: display 8 banner text checks in consent test UI
Shows: Impressum link ✓/✗, DSE link ✓/✗, plus violation cards for
wrong DSE consent wording, pre-ticked checkboxes, dark patterns,
missing reject button, no settings re-access.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 15:38:07 +02:00
Benjamin Admin 0a6ec9235e feat: 8 cookie banner legal checks (Playwright)
1. Impressum link accessible from banner (§5 TMG, LG Rostock)
2. DSE link in banner (Art. 13 DSGVO, informierte Einwilligung)
3. Wrong wording: "Zustimmung zur DSE" — DSE is Art. 13 obligation,
   not consent. Correct: "zur Kenntnis genommen"
4. Reject button visible (§25 TDDDG, no hidden reject)
5. Pre-ticked checkboxes detected (EuGH C-673/17 Planet49)
6. Dark Pattern: button size comparison — accept vs reject area
   ratio >2.5x or font size ratio >1.5x = dark pattern
7. Cookie Wall detection (Phase B — site blocked after reject)
8. Re-access to settings (Art. 7(3) — revocation as easy as consent)

All checks run via Playwright on the actual rendered banner.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 11:55:54 +02:00
Benjamin Admin c5b22e0c99 fix: derive intake flags from DETECTED SERVICES, not from text content
Fundamental architecture fix: data processing happens through APIs/scripts/
cookies — NOT through visible page text. A news site about healthcare does
NOT process health data.

Before: Qwen reads website text → guesses "health_data: true" (WRONG)
After: Google Analytics detected → tracking: true (CORRECT, deterministic)

New flow: detect services from HTML → map service categories to flags →
feed flags into UCCA assessment. No LLM needed for flag extraction.

SERVICE_TO_FLAGS maps categories: tracking→tracking, marketing→marketing+
third_party_sharing, payment→payment_data, heatmap→profiling, etc.
SPECIFIC_SERVICE_FLAGS for Klarna (Art.22), Stripe (US transfer), etc.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 08:37:51 +02:00
Benjamin Admin 0f3ec9061e fix: false positive findings + restore docs-src + §312k ecommerce filter
1. Intake prompt: "BETREIBER verarbeitet" statt "Text erwaehnt".
   IHK berichtet ueber Gesundheitsdaten → false. Vorher: true.
2. §312k Check: nur bei E-Commerce/Abo-Websites (Warenkorb, Shop, PayPal etc.)
   IHK hat keine Vertraege → kein Kuendigungsbutton noetig.
3. docs-src/ restored from commit 9824304

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 08:26:59 +02:00
Benjamin Admin e318215cc5 refactor: split agent_analyze_routes (420→309 LOC) + agent docs + migration
- Extracted website compliance checks + helpers to website_compliance_checks.py
- Created agent documentation (zeroclaw/docs/compliance-agent.md)
- DB migration 086 executed (compliance_agent_scans table)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-02 08:22:52 +02:00
Benjamin Admin 6864849115 feat: Phase 11 — granular cookie category testing
Tests each consent category in isolation:
- Phase D: Only "Statistics" enabled → checks if only analytics loads
- Phase E: Only "Marketing" enabled → checks if only ads load
- Phase F: Only "Functional" enabled → checks no tracking loads

CMP-specific category selectors for Cookiebot, OneTrust, Usercentrics,
Didomi. Generic fallback via toggle/checkbox keyword detection.

SERVICE_CATEGORY_MAP maps 35+ services to expected categories.
Violations: "Facebook Pixel loads with only Statistics enabled" = miscategorization.

Frontend: category test results shown below Phase A-C with
per-category violation cards.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 21:15:23 +02:00
Benjamin Admin f6536e8d08 fix: Use Array.isArray for legalHolds check
legalHolds can be a JSONB object {} instead of an array [], so
the || [] fallback wasn't sufficient. Array.isArray handles all
edge cases (null, undefined, object, string).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 21:12:28 +02:00
Benjamin Admin e3f26d7572 fix: Defensive legalHolds check in Loeschfristen
getActiveLegalHolds() crashed with "e.legalHolds.filter is not a
function" when legalHolds was null/undefined (e.g. old DB entries
without the JSONB field). Added fallback to empty array.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 21:06:24 +02:00
Benjamin Admin a3619c10d7 fix: Re-export STEP_EXPLANATIONS from StepHeader
VVT and Loeschfristen pages imported STEP_EXPLANATIONS as a named
export from StepHeader.tsx, but it was only imported (not re-exported).
This caused "Cannot read properties of undefined (reading 'vvt')"
at runtime. Adding the re-export fixes both pages.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 20:53:38 +02:00
Benjamin Admin d880c9d098 test: E2E tests — 47/49 passing against live instance
Results (https://macmini:3007):
- sdk-module-reachability: 40/42 (loeschfristen+vvt pre-existing bugs)
- vendor-transfers: 4/4
- isms-assets: 3/3
- document-generator: 3/4 (category label mismatch)

Added: playwright-live.config.ts (no webServer, live instance testing)
Test data NOT cleaned up — profiles persist for manual review.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 19:34:37 +02:00
Benjamin Admin d3b43250b8 test: Playwright E2E tests for SDK modules (5 specs)
New E2E test specs:
- sdk-module-reachability: Tests 40+ SDK routes for 404/crash
- scope-profiling: Three customer profiles (Startup/KMU/Enterprise)
  with screenshots at each step — data NOT cleaned up
- document-generator: Template library, categories, recommendations
- vendor-transfers: Transfer tab, explanations, adequacy list
- isms-assets: Asset register tab, form, CRUD

All tests configured to run against https://macmini:3007
Screenshots saved to e2e/test-results/ for manual review

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 19:13:52 +02:00
Benjamin Admin d1fb19810b fix: Remove premature closing brace in SECTION_FIELDS
The SECTION_FIELDS object was prematurely closed before the TOM and DPA
sections, causing a build-time syntax error. Removed the extra closing
brace so TOM and DPA fields are correctly inside the object.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 15:08:23 +02:00
Benjamin Admin 062d607da9 feat: Scope questions, placeholder mappings, example contexts
Scope questions (compliance-scope-data.ts):
- 7 new questions: org_has_employees, org_has_social_media,
  org_has_video_conferencing, proc_uses_ai_tools, proc_byod_allowed,
  prod_ugc_platform, org_cert_iso27001

Template recommendations updated:
- employee_dsi/applicant_dsi now triggered by org_has_employees
- ai_usage_policy triggered by proc_uses_ai_tools
- byod_policy triggered by proc_byod_allowed (required when yes)
- social_media_dsi triggered by org_has_social_media
- video_conference_dsi triggered by org_has_video_conferencing
- community_guidelines/terms_of_use triggered by prod_ugc_platform

Placeholder mappings (contextBridge-helpers.ts):
- 30+ new mappings for: whistleblower, video conference, AI policy,
  BYOD, consent, social media, transfer/SCC, DSI fields
- SECTION_COVERS updated for template relevance detection

Example contexts: ai_usage_policy_de, employee_dsi_de,
social_media_dsi_de, tia_de

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 13:43:09 +02:00
Benjamin Admin ef8eead513 feat: Adequacy decisions, DPF check, customer guidance for transfers
New: adequacy-decisions.ts
- Complete list of 15 countries with EU adequacy decisions (Art. 45)
- EU/EEA country set (30 countries)
- getTransferRequirement() — determines SCC/TIA/certification needs
  per country code with human-readable explanations
- US special handling: DPF certification required, check URL included

Updated: transfers/page.tsx
- "Was muss ich tun?" explanation section with 3 options:
  1. Adequacy decision (green) — no action needed
  2. DPF certification (blue, US only) — check dataprivacyframework.gov
  3. SCC + TIA required (amber) — link to Document Generator
- Collapsible adequacy countries table (15 countries with restrictions)
- Schrems II background explanation for customers
- Customer guidance written for non-experts who never heard of TIA/SCC

Updated: templateRecommendations.ts
- SCC+TIA rules now consider DPF certification and adequacy status
- us_dpf_only → SCC/TIA optional (not required)
- adequate_only → SCC/TIA not recommended

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 12:57:54 +02:00
Benjamin Admin e58c96eb70 feat: Asset register in ISMS module (ISO 27001 Annex A.5.9)
New "Assets" tab in the ISMS module for information asset management:
- CRUD for information assets (hardware, software, data, services,
  people, facilities)
- CIA protection need matrix (confidentiality, integrity, availability)
  with normal/high/very_high levels
- Information classification (public, internal, confidential,
  strictly confidential) with color-coded badges
- Category filter (all/hardware/software/data/service/people/facility)
- Stats cards (total, by category, high protection need count)
- CSV export for ISO 27001 audits
- Edit/delete per asset
- localStorage persistence (same pattern as compliance_scope)

Types: InformationAsset, AssetCategory, AssetClassification,
ProtectionLevel interfaces + label/color maps

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 12:32:24 +02:00
Benjamin Admin 03c17987a1 feat: Third-country transfer tab in Vendor Compliance module
New "Drittlandtransfers" tab in the Vendor Compliance sidebar:
- Aggregates all vendor processing locations with non-EU countries
- Traffic light system: green (EU/adequacy), yellow (SCC exists),
  red (no transfer mechanism)
- Stats cards: total, EU+adequate, third-country, action required
- Filter by status (all/OK/review/action required)
- Table with vendor name, country, mechanism, SCC status, TIA status
- "TIA erstellen" link to Document Generator for third-country vendors
- Help text explaining Schrems II / Art. 46 DSGVO requirements

Uses existing data model — no new API endpoints or DB tables needed:
- vendor_vendors.processingLocations (isEU, isAdequate)
- vendor_vendors.transferMechanisms
- vendor_contracts.documentType = 'SCC'

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 11:16:19 +02:00
Benjamin Admin 9f4c4abb84 feat: Document recommendation UI in generator
New RecommendedDocuments component shown above the template library:
- Evaluates scope answers + compliance level (L1-L4)
- Groups templates into required/recommended/optional
- Shows profile label (Startup/KMU/Extended/Enterprise)
- Cards link to actual templates — click opens in generator
- Optional section collapsed by default
- Only visible when scope has been completed

Renders as purple gradient panel with grid cards, each showing
template name and availability status.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 11:06:56 +02:00
Benjamin Admin d942b21354 feat: SCC + TIA templates for third-country transfers
New templates for the Vendor Compliance module:
- 105: Transfer Impact Assessment (TIA) — Schrems II risk assessment
  with country evaluation, government access assessment, supplementary
  measures, risk matrix, and go/conditional/deny decision
- 105: SCC Companion Document — annexes to EU Decision 2021/914
  (module selection C2C/C2P/P2P/P2C, party details, data description,
  TOMs, sub-processor list)

Template recommendations: SCC+TIA triggered by tech_third_country answer
Generator: New "Drittlandtransfer" category

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 10:19:56 +02:00
Benjamin Admin 4ff6050f43 feat: Template recommendation engine — bridges scope to document generator
Fixes critical gap: 50+ templates were unreachable because the Compliance
Scope Engine only outputs 23 document types, while the database has 70+.

New: templateRecommendations.ts
- 25 template rules that map scope answers to specific templates
- Covers ALL previously orphaned templates (HR-DSI, whistleblower,
  AI policy, BYOD, security policies, community guidelines, etc.)
- Each rule evaluates scope answers + compliance level to determine
  required/recommended/optional status
- Key triggers:
  - employee_count > 0 → employee_dsi, applicant_dsi
  - employee_count >= 50 → whistleblower_policy (HinSchG Pflicht!)
  - ai_usage != none → ai_usage_policy
  - business_model = platform → community_guidelines, terms_of_use
  - cert_target = iso27001 → isms_manual
  - webshop = yes → widerruf

Updated: scopeDefaults.ts
- getRecommendedDocuments() expanded with all 60+ document types
- L1→L4 graduated recommendation (required/recommended/optional)

Updated: _constants.ts
- Consolidated AI governance into internal_policies category

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 10:12:15 +02:00
Benjamin Admin 42e02fe72d feat: Phase 6 — Integration + QS (categories, scope defaults, examples)
Phase 6 of the Document Templates Masterplan:

- Categories: Consolidated AI governance into internal_policies,
  removed redundant category
- scopeDefaults.ts: Added getRecommendedDocuments() function that
  maps L1-L4 compliance levels to required/recommended/optional
  document types (~60 types across 4 tiers)
- Examples: Added dpa_de.json, tom_de.json, whistleblower_de.json
  example contexts for the document generator

Document recommendation per level:
- L1 (Startup): 5 required (DSI, Impressum, AGB, Cookie)
- L2 (KMU): +6 recommended (AVV, TOM, VVT, Löschkonzept, etc.)
- L3 (Extended): +16 recommended (Security concepts, policies, HR DSI)
- L4 (Enterprise): +25 recommended (ISMS, BCM, all policies)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 09:36:48 +02:00
Benjamin Admin 3984f39329 feat: Phase 5 — Special templates (AI policy, BYOD, ISMS, consent, video DSI)
Phase 5 of the Document Templates Masterplan:

- 104: 5 new special templates:
  - ai_usage_policy: AI usage policy (AI Act Art. 4 training obligation,
    forbidden inputs, quality check, labeling, TDM opt-out)
  - byod_policy: Bring Your Own Device (container solution, remote wipe,
    DSFA, cost sharing options)
  - consent_texts: Double-Opt-In texts, newsletter, marketing, tracking,
    profiling consent, unsubscribe confirmation
  - video_conference_dsi: Video conference privacy notice (Zoom/Teams/Meet,
    recording consent, third-country transfer)
  - isms_manual: ISMS handbook (ISO 27001, document structure map to all
    other templates, PDCA cycle, management review)

Generator: 6 new categories (AI governance, ISMS, consent, special DSI,
internal policies)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 09:25:32 +02:00
Benjamin Admin 4417938558 feat: Phase 3 — Security + HR/Vendor/BCM policies
Phase 3 of the Document Templates Masterplan:

- 103: 4 new security policies (information_security_policy, password_policy,
  encryption_policy, access_control_policy) + updates for CRA (056) and
  all 15 HR/Vendor/BCM policies (072)

New templates:
- Information Security Policy: ISMS-Leitlinie (ISO 27001, BSI, NIS2)
- Password Policy: BSI/NIST compliant (12+ chars, MFA, no forced rotation)
- Encryption Policy: BSI TR-02102, algorithms, key management, TLS config
- Access Control Policy: RBAC, Least Privilege, Zero Trust, rezertification

Updates: AI Act + NIS2UmsuCG references for CRA and all 15 HR/Vendor/BCM
Generator: 6 new categories (security, HR, data, vendor, BCM policies)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 09:05:03 +02:00
Benjamin Admin 90c7f02b40 feat: Phase 2 — Security Concepts + DSFA + DSR updates
Phase 2 of the Document Templates Masterplan:

- 101: Security Concepts v2 (7 templates) — NIS2UmsuCG references,
  BSI Grundschutz++ modernization, AI Act cross-references,
  Zero Trust principle, ransomware-protected backups, NIS2 logging
- 102: DSFA + Pflichtenregister + DSR v2 — AI Act Art. 9 for DSFA,
  NIS2UmsuCG for Pflichtenregister, tenant_id fix for DSR processes

All 16 templates reviewed — already at good product level, only
incremental updates needed (standards references, cross-doc links).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 08:45:04 +02:00
Benjamin Admin f591871277 feat: Phase 1 — Whistleblower + Cookie/Impressum + HR-DSI templates
Phase 1 of the Document Templates Masterplan:

- 098: Whistleblower-Richtlinie (HinSchG) — 10 sections, anonymous
  reporting, 7-day confirmation, 3-month feedback, reprisal protection
- 099: Cookie-Banner + Impressum updates — OS-Plattform discontinued
  note (July 2025), description updates
- 100: Applicant DSI + Employee DSI — two new HR privacy notices with
  § 26 BDSG, 6-month retention (applicants), modular blocks for video
  interviews, talent pool, IT monitoring, company vehicles, works council

Generator: 25 new fields (whistleblower, applicant, employee categories)
Categories: whistleblower, hr_dsi added to document generator

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 08:29:52 +02:00
Benjamin Admin bae59e2ce0 feat: Document Templates v2 — 11 migrations + scope-based generator
Complete overhaul of document generator templates based on paragraph-by-paragraph
legal review of attorney-drafted templates (TOM, AVV, AGB, DSI, Community
Guidelines, Nutzungsbedingungen, Widerrufsbelehrung, Cookie-Richtlinie).

Templates (11 migrations 087-097):
- 087: TOM-Dokumentation v2 (11 categories incl. Trennungskontrolle)
- 088: AVV Art. 28 DSGVO (complete, §§ 1-11, 3 annexes)
- 089: Cross-document updates (Löschkonzept DIN 66399, VVT recipients)
- 090: AGB SaaS/Shop v2 (18 §§, B2B/B2C, IoT, physical goods, IP protection)
- 091: Community Guidelines v2 (3 tones, 11 modular categories, DSA-compliant)
- 092: Media & Content modules (MStV, AI Act Art. 50, UWG, Pressekodex)
- 093: DSI/Privacy Policy v2 (Art. 13 complete, shop+corporate modules)
- 094: Nutzungsbedingungen (Terms of Use, UGC, tipping, wallet, CC licenses)
- 095: Widerrufsbelehrung (SaaS + physical + IoT bundle + combo)
- 096: Social Media DSI (Facebook, YouTube, LinkedIn, TikTok, Meta Pixel)
- 097: Cookie-Richtlinie v2 (TDDDG § 25, consent banner, browser links)

Frontend (generator):
- scopeDefaults.ts: L1-L4 scope-based defaults from Compliance Scope Engine
- contextBridge.ts: TOMCtx + DPACtx interfaces (70+ new fields)
- contextBridge-helpers.ts: 35+ placeholder mappings for TOM/DPA/AGB
- _constants.ts: 120+ new generator fields (TOM, DPA, AGB, community,
  media, social, nutzungsbedingungen, widerruf, cookie, shop, IoT)
- page.tsx: Auto-prefill TOM/DPA from scope engine decision

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-01 01:18:33 +02:00
Benjamin Admin 58957a4aaa fix: Playwright user permission + etracker DSE matching + CMP skip
1. Dockerfile: install Playwright AS appuser (not root) so chromium
   binary is accessible at runtime. Was causing 500 error.
2. DSE service matching: text-search fallback when LLM extraction fails.
   If "etracker" appears in DSE text, mark as documented even without
   LLM parsing the service list.
3. CMP skip: consent managers in category "cmp" skipped (not just "other"
   with id "cmp").

NOT DEPLOYED — RAG pipeline is running.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 19:36:46 +02:00
Benjamin Admin cedc5de15d feat: Phase 10 — Playwright website scanner replaces httpx
New /website-scan endpoint in consent-tester service:
- Real browser renders JavaScript (finds dynamic content)
- Clicks navigation menus (discovers hidden sub-pages like IHK DSB page)
- Follows links within DSE to find regional privacy policies
- Collects rendered HTML for each page (after JS execution)

Backend integration:
- agent_scan_routes tries Playwright first, falls back to httpx
- DSE text and HTML extracted from Playwright-rendered pages
- Service detection runs on rendered HTML (catches JS-loaded scripts)

Also fixes:
- GA regex: G-[A-Z0-9]{8,12} prevents CSS class false positives
- etracker added to service registry
- External page scanning blocked (same-domain only)
- CSS/JS/image files excluded from page list

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 19:16:50 +02:00
Benjamin Admin 5eeef3a9c3 fix: 4 bugs from IHK scan — false positives + missing etracker
1. GA regex: G-\w{5,} matched CSS classes (g-7031048). Now requires
   G-[A-Z0-9]{8,12} (uppercase after G-, 8-12 chars = real GA4 ID)
2. External page scanning: DSE-internal links now SAME DOMAIN only.
   Previously followed links to etracker.com, google.de/policies etc.
   and detected services on THOSE sites as IHK services.
3. Added etracker to service registry (DE, ePrivacy-certified)
4. CSS/JS/image files excluded from page scanning
5. Navigation-pattern links for deeper DSE sub-pages

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 19:08:07 +02:00
Benjamin Admin 891fc5bea0 docs: add keyword-based checker problem to migration instruction
mandatory_content_checker.py keywords break with alternative formulations.
Solution: LLM-based check per mandatory field (9 calls, parallelizable).
For other session to implement alongside Dict→Control migration.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 18:18:45 +02:00
Benjamin Admin fff47cc52e fix: 4 bugs from IHK Konstanz scan validation
1. DSE-Matcher: Google/YouTube false match — now requires 2+ word match
   for provider-name fallback, not just "Google" matching YouTube section
2. AGB/Widerrufsbelehrung: only_ecommerce flag — skips for non-shop
   websites (detected via payment providers, cart keywords)
3. DSE-internal link following — scanner now discovers links WITHIN the
   privacy policy and scans those too (finds regional DSE sub-pages)
4. Expanded keyword synonyms for DSE mandatory checks:
   - "Zweck und Rechtsgrundlage" now matches "zwecke"
   - "behoerdlichen datenschutzbeauftragt" matches DSB
   - "aufsichtsbehörde" with umlaut matches

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 17:57:19 +02:00
Benjamin Admin 0f3ba9c207 test: Lit-Mapping validation — Dict vs Control Library comparison
8 test cases with deliberately wrong legal basis assignments:
- Cookie tracking on lit. f (should be lit. a)
- Analytics on lit. b (should be lit. a)
- Newsletter on lit. f (should be lit. a)
- Klarna without Art. 22
- Session recording on lit. f
- 2 correct cases (should NOT trigger findings)

Runs both hardcoded dict AND Control Library query, compares results.
If Control Library passes all → dict can be removed.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 16:56:38 +02:00
Benjamin Admin b53b36fdc5 feat: 5-tab agent UI — PDF export, compare, auth test, all proxies
- 5 tabs: Schnellanalyse, Website-Scan, Cookie-Test, Vergleich, Login-Test
- PDF download button in ScanResult
- CompareResult: side-by-side compliance comparison table
- AuthTestResult: 5 post-login checks with legal refs
- API proxies: /scans/pdf, /compare, /authenticated-scan
- Compare: textarea for 2-5 URLs, parallel scanning

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 16:43:08 +02:00
Benjamin Admin 2c9cea74e3 docs: instruction for hardcoded knowledge → Control Library migration
6 files with hardcoded legal knowledge identified. Review deadline 2026-07-01.
legal_basis_validator.py marked with warning log on every use.
Instruction file for other session to execute migration.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 16:33:48 +02:00
Benjamin Admin 85c4cbbf37 fix: increase scan proxy timeout from 3 to 5 minutes 2026-04-29 16:24:22 +02:00
Benjamin Admin 4bf92f42b8 feat: Phase 9 — Authenticated Testing + Legal Basis Validator (lit. mapping)
Phase 9: Playwright login + 5 post-login checks:
- §312k BGB: Kündigungsbutton (2 Klicks)
- Art. 17 DSGVO: Konto löschen
- Art. 20 DSGVO: Daten exportieren
- Art. 7(3): Einwilligungen widerrufen
- Art. 15: Profildaten einsehen
Auto-detects login form selectors. Credentials destroyed after test.

Legal Basis Validator: Checks 7 common lit-mapping mistakes:
- Cookie tracking on lit. f instead of lit. a (Planet49)
- Analytics on lit. b (contract overextension)
- Klarna without Art. 22 reference
- Session recording without consent
Integrated into website scan pipeline.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 16:08:41 +02:00
Benjamin Admin 8336c01c5c feat: Phase 6-8 — PDF export, recurring scans, multi-website compare
Phase 6: PDF export via WeasyPrint — POST /agent/scans/pdf generates
printable compliance report with findings table, service comparison,
risk badge, and legal disclaimer.

Phase 7: Recurring scans — POST /agent/monitored-urls to add URLs,
POST /agent/run-scheduled triggers all enabled scans (cron/ZeroClaw).
In-memory storage with DB upgrade path.

Phase 8: Multi-website compare — POST /agent/compare with 2-5 URLs,
parallel scanning, comparison table (risk, findings, services, compliance
features per site).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 15:27:51 +02:00
Benjamin Admin e35db90232 feat: Phase 5 — DB persistence for scan results + Phase 10 in plan
- Migration 086: compliance_agent_scans table (findings, services, corrections)
- agent_history_routes.py: POST /scans (save), GET /scans (list), GET /scans/{id}
- Scan results survive page reloads and can be reviewed later
- Phase 10 (Playwright website scanner) added to product roadmap

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 15:17:51 +02:00
Benjamin Admin 53774886e7 perf: Phase 4 — parallel page fetching (asyncio.gather)
Scan pages in parallel instead of sequential. Reduces scan time
from ~10s (5 pages × 2s) to ~3s (all pages at once).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 15:09:03 +02:00
Benjamin Admin 5c5054f740 feat: Phase 3 — registry 82 services, mandatory checker, SDK flow step
- website_scanner.py: imports from master service_registry.py (82 services)
- agent_scan_routes.py: mandatory content checks (documents + DSE sections)
- steps-betrieb.ts: Compliance Agent step added to SDK Flow (seq 5000)
- PLAN: Phase 9 (Authenticated Testing) added to product roadmap

Mandatory checks know what MUST be there:
- Documents: Impressum, DSE, AGB, Widerrufsbelehrung
- DSE content: 9 Art. 13 DSGVO fields (DSB, Speicherdauer, etc.)
- Impressum content: 5 §5 TMG fields (GF, HRB, USt-ID, etc.)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 15:04:44 +02:00
Benjamin Admin 642382cbe8 feat: Mandatory Content Checker — knows what MUST be there
Three check levels:
1. Documents: Impressum, DSE, AGB, Widerrufsbelehrung must exist as pages
2. DSE content: 9 Art. 13 DSGVO mandatory sections (Verantwortlicher,
   DSB-Kontakt, Zwecke, Rechtsgrundlagen, Speicherdauer, Betroffenenrechte,
   Beschwerderecht, Drittlandtransfer, Profiling)
3. Impressum content: 5 §5 TMG mandatory fields (GF, Handelsregister,
   USt-ID, Anschrift, Kontakt)

Detects both missing documents AND missing content within documents.
Also catches HTTP errors (page exists but returns 404/500).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 14:23:22 +02:00
Benjamin Admin f219b9c244 feat: Master Service Registry — 82 third-party services across 15 categories
Tracking (12), Marketing/Ads (9), Newsletter (8), CDN/Fonts (7),
Chatbots/Support (7), Payment (5), Heatmaps (4), A/B Testing (3),
Tag Managers (3), Push (3), Video (4), Social (3), Error Tracking (4),
CRM (3), Maps (3), Captcha (3), Accessibility (2), CMP (1).

Each entry: regex, provider, country, EU adequacy, consent requirement,
legal reference. Pure data, no logic.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-29 14:21:32 +02:00