[Rspamd-Users] How to you compine spamtrap, training and avoiding new

[Christian]

Hello all,

I am interested in how you use spamtraps and if you have found ways to identify new spam emails via a spam trap before the same spam emails are delivered to users.

Here is an example:
In the respective Spam waves I recognize again and again 1:1 identical Spam e-mails or strongly similarly structured e-mails. My goal is to learn these as spam as early as possible via the spam trap, so that further delivery attempts to real users already lead to a classification of the emails as spam.

I use a spamtrap (a whole domain with catchall) for this purpose. Incoming emails there are automatically learned as spam (Fuzzy & Bayes). For all other mailboxes I have defined a low greylisting threshold. So far this configuration shows only a small improvement.

Is there a better way to do this or how do you use the spamtrap?

Greetings Christian