The training register incl. scan attachments and checksums is now part
of the backup export/import — the Art. 4 evidence chain survives device
moves. Older backups (v1/v2) remain importable.
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- Manual participant entry for paper tests: name + correct answers →
register entry with source 'paper' (app runs are tagged 'app')
- Scan upload per register entry: marked paper sheets (image/PDF,
max 1 MB) stored locally as data URL and anchored with a SHA-256
checksum of the original file; view scan in-app, checksum shown
- Document verification: certificate and training register printouts
carry a SHA-256 checksum footer over the document content — a
presented printout can be re-verified against the tool; certificate
additionally references its register entry ID (evidence chain
certificate ↔ register ↔ scan)
- Register printout extended with source, evidence checksum and ID
columns; record deletion with confirmation warning
- UI labels in all nine languages; 73 tests green
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
The answer key is self-contained for markers: each entry shows the
question text, the correct answer spelled out (letter + wording) and
the explanation — no need to hold the test sheet next to it.
Test updated accordingly (72 green).
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- Paper version of the AI literacy test: name/date/department fields,
12 questions with tick boxes (A/B/C), marking line and signature
block — for employees without computer access; contains no answers
- Answer key for markers: correct letter plus explanation per question,
pass threshold noted
- Training handout: the 12 core rules (question + explanation) as
Art. 4 training material to distribute before the test
- All printable with company letterhead via the existing print/PDF
pipeline and downloadable; buttons in the quiz view under 'Print
training materials'; UI labels in all nine languages (document body
DE/EN with EN fallback, full localization follows the content pack)
- 3 new tests (72 total)
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
All 12 AI literacy test questions (question, 3 options, explanation —
540 strings) translated to FR/IT/ES/PL/CS/SK/RO with official EU AI
Act terminology and locale-appropriate GDPR abbreviations (RGPD/RODO/
GDPR); light humour of the distractor answers preserved.
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- Simple 12-question basics quiz ('for dummies'): how AI works,
hallucinations, confidential data, human oversight, chatbot
disclosure (Art. 50), prohibited practices (Art. 5), high-risk
(Annex III), GDPR, internal incident reporting (Art. 73), shadow AI
approval, content marking, the Art. 4 duty itself — 3 options each,
pass threshold 70%
- Evaluation view with per-question explanations (learning effect),
pass/fail stamp badge
- Printable certificate (letterhead) as Art. 4 evidence: name, date,
score, topics covered, signature lines — files into the existing
training plan documentation
- Local training register (localStorage): every run recorded (name,
date, score, status), printable as continuous Art. 4 evidence
register; cleared with all-data delete
- New nav entry; UI strings in all nine languages; question bank
currently DE/EN (translations for the other seven languages follow
in the next commit)
- 6 new tests (69 total)
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- New RFC 5545 generator: all future obligation deadlines (per system)
plus upcoming AI Act application dates as all-day events, each with
two DISPLAY alarms (30 and 7 days ahead), stable UIDs for
re-import/update, proper escaping and 75-octet line folding, CRLF
- Export button in the tasks view downloads ai-act-fristen-<date>.ics
(text/calendar) — opens in Outlook, Apple Calendar, Google Calendar,
Thunderbird on phone and desktop; no server involved (local-first)
- Event titles and descriptions localized via the active content pack
(all nine languages); button and hint translated in all dictionaries
- 5 new tests: calendar frame, future-only filter, stable UIDs, alarm
count, escaping/folding, localized summaries (63 total)
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
Complete translations (UI, content pack, obligations catalogue,
national modules, classification reasons, document strings, Annex IV
sections incl. new FR/IT/ES section translations, procurement spec,
conformity declaration, all 10 template documents) for pl/cs/sk/ro,
using the official terminology of the respective language versions of
Regulation (EU) 2024/1689 (e.g. podmiot stosujący, zavádějící subjekt,
nasadzujúci subjekt, implementator). Locale architecture extended to
nine languages with EN fallback reserved for future additions;
language switcher, browser detection, locale-aware date formats
(dots for de/pl/cs/sk/ro), localized Annex IV title; language hints
updated; dictionary-completeness and real-translation tests now cover
all seven non-DE/EN locales (58 tests green).
Note: pl/cs/sk/ro legal texts are machine-drafted against official
terminology — native legal review recommended before market launch.
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
Every user-visible text is now fully available in all five languages
(DE/EN/FR/IT/ES) — no English fallback remains in the product:
- Content pack: AI definition criteria, prohibited practices, Annex III
areas, transparency triggers, deadlines (36 entries)
- Obligations catalogue: all 26 EU obligations (title + description)
- National modules: all 6 countries incl. authority descriptions
- Classification reasons (classify engine)
- Document strings, risk/role/status labels, disclaimers, Annex IV
- Procurement spec (12 requirements, DOIT/DEVE/DEBE priorities) and
EU declaration of conformity (numbering-preserving field prefill)
- All 10 template documents (works council, employee info, FRIA,
incident report, AI policy, training plan, works agreement,
reporting policy, approval request, audit checklist)
Official terminology of the FR/IT/ES language versions of Regulation
(EU) 2024/1689 (e.g. hypertrucages/ultrafalsificaciones, contrôle
humain, alfabetizzazione in materia di IA); German statute names
(BetrVG, HinSchG, KI-MIG) kept as proper nouns.
Language hints updated; tests now assert real translations instead of
EN fallback (58 tests green).
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- Full UI dictionaries for French, Italian and Spanish (all views,
wizard, legal gate, templates); browser-language auto-detection
extended to all five locales
- Fallback architecture: LText now carries optional fr/it/es entries;
regulatory legal texts (content pack, classify reasons, document
strings) fall back to English until legal translations land —
clearly stated in the language setting
- Locale-aware date formatting (DE dots, FR/IT/ES slashes, EN ISO)
- PWA: web app manifest, network-first service worker with offline
cache fallback, generated icons (512/192/apple-touch) — installable
on phone home screen and macOS dock
- 2 new tests incl. dictionary-completeness check for FR/IT/ES
(58 total)
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- Company-wide templates view: internal AI policy (framework), AI
literacy training plan & record (Art. 4) listing inventory systems,
works agreement on AI (Sec. 87(1) No. 6 / Sec. 90 BetrVG draft)
- System-specific forms in detail view (role/risk/country gated):
works council information letter (Sec. 90 BetrVG, DE module),
employee information on AI use (Art. 26(7) / Law 132/2025),
FRIA working template (Art. 27), serious incident report form
(Art. 73 incl. 15/10/2-day deadlines)
- All templates bilingual (DE/EN), letterhead-printable, placeholder-
based drafts with legal disclaimer
- 8 new tests (52 total)
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- Full i18n (DE/EN): UI, rules engine, content pack and generated
documents; language switcher, browser-language default, architecture
extensible to further EU languages via localized source texts
- National country modules (opt-in per company profile): DE (KI-MIG/
BNetzA supervision, works council co-determination Sec. 87 BetrVG,
GDPR/BDSG), AT (RTR AI service desk), IT (Law 132/2025: employer info
duty, deepfake criminal liability, sector decrees), ES (AESIA,
sandbox RD 817/2023), FR (CNIL guidance), NL (algorithm register);
merged into obligation derivation, checklists and documents
- New EU obligations: corrective actions (Art. 20/21), authorised
representative (Art. 22), EU database registration for public-body
deployers (Art. 49(3)), post-market monitoring (Art. 72), serious
incident reporting (Art. 73), GDPR DPIA (Art. 35 GDPR / 26(9))
- New documents: AI procurement requirements specification (MUST/SHOULD
matrix for vendors) and EU declaration of conformity draft (Annex V)
- Cross-system checkable task list (due now / upcoming / done) with
progress bar, one-tap toggling, per-system links
- Legal safety: first-run acknowledgement gate (working aid, no legal
services), legal notice panel in settings, result-screen disclaimer,
disclaimers on every document
- Responsive layout for phone/tablet (top-bar nav, scrollable tables,
larger tap targets)
- 12 new tests (44 total): EN pack parity, national merge, checklist
grouping, new document generators
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
- Full redesign: paper tone, serif typography, hairlines, rectangular
stamp badges, mono labels — technical print look instead of SaaS style,
no gradients/shadows/emoji
- Company profile settings (name, address, contact person, email, phone,
website, logo upload as local data URL) persisted in localStorage and
included in JSON backup (schema v2, v1 imports still supported)
- Print/PDF output via letterhead print view (window.print): logo,
company name and contact data in header, contact line in footer;
applies to risk report, action plan, Annex IV skeleton and inventory
- Company block also embedded in generated Markdown headers
- Minimal auditable Markdown renderer (no external parser), HTML-escaped
- 8 new tests (32 total)
Claude-Session: https://claude.ai/code/session_017YjohVNx1ZGNM4MX7tHpfv
