Can a local AI pentesting agent produce findings I can independently reproduce?
An overnight Strix experiment with local models, a vulnerable web app, and a pentest report that yields results.
AI RED TEAMING NOTE 001 · AI SECURITY
Building an AI For a Spam Classifier (With Python)
A study note on how to implement a spam classifier with AI from scratch using python.
Read field log 10 min read
SPAM CLASSIFICATION● CHECKING SPAM
01 / DATAspam.csvget data, normalize and standarize it
→
02 / MODELtrainingapply naive bayes and learn what is spam and what isn't
→
03 / USEevaluatetest model against real world spam
$ wget https://example.com/spam.csv$ ./feature-extraction -i spam.csv -o spam_clean.csv[done] 153784 features identified$ ./train_model -i spam_clean.csv$ ./check_spam "You won 1'000'000.- $!!"[DANGER] SPAM DETECTED!
LATEST TRANSMISSIONS
Full archive Read the field notes.
DVNLL LABS / VIDEO WORKBENCHWATCH THE BUILD
This Small Web Bug Led to Full Server Takeover
What happens when a vulnerability that looks small at first glance becomes the first step toward full server compromise? In this episode, I walk through the attack path I found while hacking the retired Hack The Box machine Cap.
Watch videoOne useful transmission.
No content treadmill.
The email digest is being wired up. Until then, the RSS feed is live and contains every new field note.
Portable feed · full archive · no platform lock-in