[Netmail-support] ? re: anti-spam agent

[Benjamin A. <be...@cu...>]

Hi
RedHat 2.1
Netmail 3.52a
SpamAssassin 2.64

Part 1
I am trying to modify the anti-spam agent to report on all messages.  I 
have no experience with spamassassin or nmap so I will be learning as I 
go.  I have taken some code from the original spamd.raw and have it 
logging the spam score to syslog.  I would like to have it also report 
the score in the message so I don't have to log into my mail server.  I 
was first setting the X-SPAM-FLAG value to YES or NO depending on 
whether it is spam or not.  Then I will figure out how to insert the 
spam score.  The flag piece works fine but what I found was during the 
second check when it looks for SpamAssassin in the message and then 
executes the QDONE command it seems to return the local email address 
instead of the QDONE return code of 1000 causing the return test to 
fail.  I guess it doesn't really matter because it is exiting anyways 
but it logs it in maillog.  What I have seen the first time through 
when it pushes values into the envelope they are

F<sending email address>
A<dont know what these numbers are>
X1
L<local email address>

The second time through when it reads the lines looking for 
SpamAssassin it looks like it is

F<sending email address>
A<dont know what these numbers are>
X1
SpamAssassin
L<local email address>

I guess what all this blabbering means is I am looking for info on the 
message format.  What is each line supposed to look like?  How was 
SpamAssassin inserted before the L<email> value when it comes after the 
envelope loop.  And what ever you might think is relevant but that I 
don't know enough to ask about.

Part 2.
It's much shorter.
All my spam scores are 0.0
How can I make it score something so I know SpamAssassin is really 
working?

Hope this all makes sense.
Thanks
Ben