¿Filtro de spam basado en palabras clave del contenido del mensaje?

We are. These things come up as flagged; it’s just that it’s unique (and annoying) enough that I was hoping for a way to filter it straight to trash.