Mutt, Spamassassin and AWL

As any spamassassin user would agree, the spammers aren’t dumb either. They update the characteristics of their email to avoid being filtered out. Some of those emails sneak through before you’ve updated your scores. When this happens, if you’ve enabled auto-whitelist, there are chances that the sender (though spoofed) may end up getting a negative AWL score meaning that the next time the overall score is bound to be lower, and the chances of getting the same spam message is not insignificant. I’ve seen this quite frequently with the misses. I thus have to update my scoring to catch the spam and remove the sender from the auto-whitelist. Here is a tiny mutt macro that does the latter -

macro index,pager wa "<pipe-message>grep \"\^From:\" | head -n 1 | cut -d \"<\" -f2 | cut -d \">\" -f1 | (read a; spamassassin --remove-addr-from-whitelist=\$a)<enter>" "Remove from spamassassin AWL"

Here are some other spamassassin related macros I use -

macro index,pager wc "<pipe-message>spamassassin -t | tail -n 30<enter>" "Run through spamassassin and check score"
macro index,pager ws "<pipe-message>sa-learn --showdots --spam<enter>" "Learn message as spam"
macro index,pager wh "<pipe-message>sa-learn --showdots --ham<enter>" "Learn message as ham"

Permalink | Trackback | Hacks