[Rspamd-Users] all learn conditions denied learning spam in default classifier

[Tino Hendricks]

Dear list,

I’m still trying to get my head around this cryptic message. 
Neither google nor

debug_modules = ["bayes" ,"lua_bayes", "task“]

is of much help (for me).

Today I received a mail scoring like this:

BAYES_SPAM (2.296376) [90.50%]
RCVD_DKIM_ARC_DNSWL_HI (-1)
RBL_MAILSPIKE_BAD (1) [46.23.XXX.XXX:from]
RCVD_IN_DNSWL_HI (-0.5) [46.23.XXX.XXX:from]
R_DKIM_ALLOW (-0.2) [xxxxxxxxx.net:s=root]
BAD_REP_POLICIES (0.1)
MIME_GOOD (-0.1) [multipart/alternative,text/plain]
FROM_EQ_ENVFROM (0)
MIME_TRACE (0) [0:+,1:+,2:~]
TO_DN_ALL (0)
FROM_HAS_DN (0)
MID_RHS_MATCH_FROMTLD (0)
TO_MATCH_ENVRCPT_ALL (0)
ARC_NA (0)
DMARC_POLICY_ALLOW (0) [xxxxxxxxx.net,reject]
RCPT_COUNT_ONE (0) [1]
RCVD_COUNT_ZERO (0) [0]
R_SPF_ALLOW (0) [+ip4:46.23.XXX.XXX/24]
NEURAL_HAM (0) [-0.992]
DKIM_TRACE (0) [xxxxxxxxx.net:+]
ASN (0) [asn:213373, ipnet:46.23.XXX.XXX/24, country:SC]

resulting in a total of 1.59.
Immediately I fed the mail to the „Learn“ section of the web interface which gives me the "all learn conditions denied learning spam in default classifier“.

What does it mean and what can I do to make rspamd learn that this is a bad mail?

Thank you very much.

Tino