Democratizing mobile app security; building https://xitter.cf/t.co/jTAyyWNqWk

🌎
Joined July 2018
_bhamza retweeted
In 2026, no program is closed source. We decompiled a propAMM into Rust that recompiles to the exact on-chain binary, then used it to find a critical bug that put $6.1M at risk.
22
72
15
538
30,982
This looks amazing. A cross-platform WiFi cracking tool was something deemed impossible just 10 years ago. Kudos to @derv82 for making this happen.
2
52
438
39,787
Share your feedback right inside the app scanner.
1
1
6
139
or send me a DM :_)
34
_bhamza retweeted
1/6 🧵 Another thread -- and time to put on my slightly boring “professor” hat 🤓 We’ve just released the first three chapters of The Decompilation Book! We call it a book, although technically, it’s a series of blog posts: decompilation.education/
6
64
350
15,304
still unsure if it was wayback machine, a random crawler or a random hacker poking at my scanner. xitter.cf/_bhamza/status/2097763…
I woke up to a Sentry alert timestamped at 4:52 a.m: “Input should be a valid UUID...”. It was coming from a route used to view a sample public report, which expected the UUID of a scan. Nothing critical, but it reminded me of something.
1
7
214
I woke up to a Sentry alert timestamped at 4:52 a.m: “Input should be a valid UUID...”. It was coming from a route used to view a sample public report, which expected the UUID of a scan. Nothing critical, but it reminded me of something.
1
1
4
337
Each rescan generated a new UUID, which also meant a new public report URL. The frontend works fine, but from an SEO perspective, this is terrible. A search engine/crawler sees a disappearing URL (the old scan is gone) and a completely new one in its place.
1
1
31
The fix was simple: give each sample a permanent slug instead of using the scan UUID in the public URL. Funny how a completely harmless crawler error at 4:52 in the morning accidentally reminded me of an SEO flaw. See for yourself (no signup needed): atomicsecurity.app/examples
1
20
_bhamza retweeted
Last week we released Virtual//Attack An AtomicRedTeam-style collection of 80+ attack techniques against VMware environments including ESXi, vCenter, and many more. Optimised for adversary emulation, with TA tags and TI refs Check it out 👇 github.com/ReversecLabs/virt…
1
53
2
237
10,844
Replying to @jun_song
I've seen this "no need to code" argument a lot online now with AI. At first I was trying to figure out whether this was true or not. However, in a recent venture of mine I've been working with a non-technical client. He discovered Codex, and he started creating what he needs by himself. I told him give it to me, and I will check to see whether we can use it. To say the least, it was a buggy mess, while he thought he could just deploy this to prod. I think we still need people with domain knowledge, what I produce with AI and what non-tech people produce with AI differs vastly in quality.
1
3
522
“Does it use AI/LLM?” is a recurring question I get asked about Atomic Scanner. Currently, the answer is NO. Not because I’m in denial, in the contrary, I do actually use AI to build and secure it. Here are six reasons why I didn't incorporate AI in the scanner (yet): 🧵
2
129
For all of this to work well later, I need the basic building blocks first. That’s what I’m building now. Next in the pipeline 🔜: dedicated detection rules for React Native apps.
1
29