CAS first indexed content scan keeps grep on the freshest indexed snapshot before falling back to extracted text.
The consolidated scan path uses blob-backed bytes as the authoritative indexed content surface.