Scanned / image-only PDFs have no text layer, so extraction currently returns nothing for them. Add an OCR fallback (e.g. Tesseract) in `src/lib/rag/extract.ts` for PDFs that yield little or no text.
Scanned / image-only PDFs have no text layer, so extraction currently returns nothing for them.
Add an OCR fallback (e.g. Tesseract) in
src/lib/rag/extract.tsfor PDFs that yield little or no text.