feat: Gill Sans Infant Tesseract model (BCER 0.071%) + font OTF via LFS

chore: track *.traineddata via LFS + Gill Sans font->model mapping

fix: gruffalo lowercase, remove dangerous so->Go word correction

fix: ocr_done state triggers review load after OCR completes; add Gruffalo word corrections

feat: prev button, step instructions all tabs, noise threshold 3.2, word corrections I->T G->s

Add persistent book page scope, safe titles, and OCR stall guards

Upload tessdata/gill_sans_infant.traineddata.otf

Integrate punctuation corrector into OCR routing and review learning

Add Smoke Signal diagnostic audit trail and bug report template

Fix Smoke Signal Surya OCR for new predictor API

Training phase: scripts 09+10, --model flag in 03, gold schema patch

Stages 9+10: Codex exporter and validation report

Stages 7+8: LLM normalisation and review workbench

test push

Upload 04_region_detector.py

Update smoke_signal/scripts/03_ocr_bakeoff.py

Create 03_ocr_bakeoff.py

Create 02_profile_pdfs.py

Upload 13 files