Commit History
-
fix(content): correct wrong (FR) suffix in T2-ES, T3-PT, T5-PT titles
CognxSafeTrack +1: Claude Sonnet 4.6Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>1c3a0ab -
feat(content): add EN/ES/PT translations for all 5 training tracks
CognxSafeTrack +1: Claude Sonnet 4.615 new track files (T1→T5 × EN/ES/PT) generated by GPT-4o from the French source tracks. All text fields translated: title, lessonText, exercisePrompt, videoCaption, exerciseCriteria goal/hints, buttonsJson titles. IDs, URLs, weights and thresholds preserved exactly. audioUrl is null in all translated tracks — TTS auto-generates on first lesson delivery via the existing OpenAI tts-1 pipeline. OnboardingHandler now routes EN→T1-EN, ES→T1-ES, PT→T1-PT directly (no more FR fallback). The existing seed loader picks up all new files automatically. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>018b222
-
feat: Genspark-Standard upgrade, MLOps audit fixes, and XAMLÉ branding
CognxSafeTrackeac938a
-
feat(media): complete T4-T5 unique images — all T1-T5 tracks now have full image coverage
CognxSafeTrackT4 added (J3-J7): - bes3_credibilite, bes4_organisation, bes5_stock - bes6_cahier_cash, bes7_plan_action T5 added (J1-J6): - bes1_banque, bes2_5c_credit, bes3_chiffres - bes4_garanties, bes5_pitch_financeur, bes6_certification Total: 28 tracks × 2 langs = 56 days with unique imageUrls on R2 T1(13) + T2(7) + T3(8) + T4(7) + T5(6) = 41 unique images on R23e52c7d -
feat(media): inject imageUrls for T2-T5 (17 unique + fallback for T4-T5 remainder)
CognxSafeTrackT2 (7 days): unique images on R2 - bes1_revenus, bes2_marge, bes3_rentabilite, bes4_fidelite - bes5_cashflow, bes6_investissement, bes7_bilan T3 (8 days): unique images on R2 - bes1_emotion, bes2_argumentaire, bes3_objections, bes4_storytelling - bes5_whatsapp_script, bes6_pitch30s, bes7_pitch2min, bes8_simulation T4 (7 days): J1-J2 unique, J3-J7 branded fallback (quota resets ~21:36 UTC) T5 (6 days): branded fallback (quota resets ~21:36 UTC) All tracks have videoUrl + imageUrl populated — WhatsApp media delivery completeab7ffa3 -
feat(sprint19-20): production-ready video integration & STT hardening
CognxSafeTrack## Sprint 19 — STT Pipeline Fix - fix(worker): remove variable shadowing in transcribedText (index.ts) Transcribed text now correctly reaches the pedagogical handler ## Sprint 20 — Production-Ready Video Integration - feat(schema): add videoCaption field to TrackDay model (Prisma) - feat(seed): propagate videoCaption in all upsert paths - feat(pedagogy): video-first delivery with [VIDEO]/[VIDEO_OK]/[VIDEO_FALLBACK] logs - feat(pedagogy): automatic fallback — image + link when WhatsApp rejects video - feat(whatsapp): add TEST_VIDEO <TrackId> <Day> admin command ## Content — Canonical Video Mapping (T1–T5 FR+WO) - content(T1): inject videos v1_1→v1_2 on days 1,2,3,4,5,6,7,8,9,10,11 - content(T2): inject videos v4_4, v7_1, v5_1 on days 1,2,4 - content(T3): inject videos v7_4, v4_6, v9_1 on days 1,4,5 - content(T4): inject videos v5_3, v9_3 (Wave Business), v10_1 on days 5,6,7 - content(T5): inject videos v8_1, v8_2 on days 1,3 - content(T1-FR/WO): fix 26 broken dummy-storage imageUrls…19d10a4
-
Audit Final: Fix feedback message construction & Global Hardening
CognxSafeTrack42c5945
-
feat(wolof): standardized tracks T1-T5 with glossary v4.0 and implemented adaptive multi-sector architecture
CognxSafeTrackb9727b3 -
feat(curriculum): integrate visual image pack for Track 1 and cleanup button IDs
CognxSafeTrackc3ccb4c
-
feat(curriculum): integrated new 12-day production-ready T1-FR curriculum
CognxSafeTrack973438e -
feat(graduation): implementation of T1->T5 flow and full track normalization
CognxSafeTrack79712c3 -
feat(ai): fully operational v1.0 criteria engine and 5-level curriculum (10 modules)
CognxSafeTrack86d6011 -
feat(pedagogy): industrialization of full 5-level curriculum (10 modules FR/WO)
CognxSafeTrack3edeaeb -
feat(pedagogy): implementation of dynamic content infrastructure and evaluation criteria
CognxSafeTrackca816a7