ETHSecurity Community

Bigger scary number better

models tend to way overhype the severity of their findings... Did the peeps behind this not bother triaging at all and accepted what the models claimed at face value?

… but all the findings were load bearing!

To benchmark OpenAI’s codex-security tool with other models and a simpler harness we just open-sourced a version based on the PI harness. From initial results it actually works very well running it using DeepSeek v4 flash and pro. We will release data soon. Would say this is probably the best open source vulnerability detection tool out there right now. https://github.com/Cecuro/open-security

Worked for arxiv 🤷🏾‍♂️

Gib PoC.

https://osintteam.blog/not-every-random-looking-number-is-secure-coldcard-hacked-for-over-130-millions-in-bitcoin-5db118a796d3?source=email-0cc8e2e3e2dc-1786778309885-newsletter.subscribeToProfile-2983bc435765------------------------f962671b_23e5_4939_afaf_10242efae559--------1097e5ec2b3c

worst AI image ive ever seen

omg.. 😂

Happens to the best of us 😄