Spam that gets through

is there a way to teach the filter as SPAM does come through to users?
nothing too comfortable - and we're not planning on adding this.
PMG uses SpamAssassin for spamdetection and you can use it's Bayes feature and train it

Question has been asked quite a few times in our forum:
https://forum.proxmox.com/threads/teach-spam.68759/
https://forum.proxmox.com/threads/s...for-bayes_00-and-bayes_05.105398/#post-453796

there are also a few users who have compiled their modifications - here is a recent example which incorporates bayes configuration (in German):
https://github.com/BJ48965156/pmg-adjustments

I hope this helps!