Spamcop Auto-reporting Plugin

[Alex Neuman]

Matt Kettler escribió:
> No, and such things are generally discouraged because it's a bad idea.
>
> SA had the feature to do auto-reporting to razor a long time ago, and it was
> removed at the request of cloudmark. Cloudmark has an *explicit* ban on users
> auto-reporting to razor based on the results of automated analysis tools like SA.
>
> The problem being that SA scores alone are not that reliable of an indicator of
> spam. If you autoreport to a system, be it spamcop, razor, etc, based on SA
> score, that system effectively becomes the equivalent of spamassassin. It will
> have all the same FPs, and all the same correct spam hits.
>   
While I wholeheartedly agree that reporting SPAM sent to legitimate 
addresses using an automated tool can be abused, I believe there could 
be a use for autoreporting: spamtraps.

Perhaps somebody here could help me out. Let's say I buy a domain 
(they're cheap anyways). I *know* I'm not going to receive e-mail for 
that domain since it's a throwaway, so I can turn it into a spamtrap (I 
could also use a subdomain, say, "crap.nkpanama.com", or an address on 
my actual domain which I *know* to be invalid... but I digress...).

I could set up a "mailertable" entry that says "anything for 
thisdomainiscrap.com gets sent to spamtastic at local".

Then the user spamtastic has a .forward file that says "forward the 
contents of the message to spamassassin --report".

Would this be feasible? Pros & Cons? Let me know what you think.