Five security reports within 20 minutes, from the same user, all scored as high as possible (and of course wrongly), broken formatting copy-pasted from an LLM, security policy and thread model completely ignored, and the two comments I so far got on closing two of these also sound like they fell straight out of a slopbot.
In other news, *definitely* going to go with a summer of bliss approach as well (as invented by daniel:// stenberg:// (npub1cm0…enqe)).
And will overhaul the security policy a bit.