I've had spam making it through SA 3.02 with the X-Antiabuse headers in the mail. Anyone have any ideas on how to prevent minus scores on those rule hits?
Care to be specific about which rule hit's you're talking about? AFAIK SA does not have any rules that score negative for X-Antiabuse headers.
In fact, the only negative scoring rules in 3.x I know of are:
ALL_TRUSTED BAYES_00-BAYES_40 USER_IN_* (whitelists of various sorts) RCVD_IN_BSP_* HABEAS_USER HASHCASH_* SPF_PASS (trivial score) SPF_HELO_PASS (trivial score)
And of course the AWL can take any score at all...
