I would be very interested in this as well. We run several mailscanner boxes here which are handed email in a round-robin fashion. We choose to scan all incoming and outgoing email for spam and viruses so simply whitelisting our internal addresses wouldn't work for me. Unfortunately some users who receive their email on their own systems (after MailScanner has already scanned them) have a forward rule that sends this email back out of our system (for instance, to a gmail account). In this case we end up scanning the same message twice.
<br><br>It would be great if MailScanner could identify a message that has already been identified as spam by looking for the "X-so-MailScanner-SpamCheck" line (one would, of course, want to use a customized line here so that a spammer couldn't just put the generic line in a message) and then just pass it along with the previous score rather than running the spam checks again. My thought is that this would be fairly safe because (1) we would only do this for messages that have already been found to be spam, and (2) because we only trust our own custom headers. To be a bit safter we could only allow this rule to be triggered if a message comes from certain IP addresses.
<br><br>Is there a general interest out there for this kind of a thing? If so, I wonder if it's something Julian would be interested in implementing.<br><br>Neal L.<br><br>