From: J Rice on
My goal:
Implement greylisting only on emails that appear spammy (ie, have an
inlined image and so on), based on header/body checks and (presumably)
using the FILTER: directive.

So far,
I have not been successful. I am experienced with greylisting using a
number of implementations (gld, gps, etc) but those use the
check_policy_access hook and don't appear able to take their input
from the FILTER directive. Or so it appears...

Is it possible to use a content filter as a greylist daemon? It seems
only content filters can be called from header or body checks, and all
the implementations of greylists I know of are policy daemons, not
content filters. If this is a possible implementation, however, I
might try to write one.

Jeff