- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
cross-posted from: https://sopuli.xyz/post/9771976
GitHub besieged by millions of malicious repositories in ongoing attack
Imagine if AI starts to pick up those malicious code as valid, and when you ask it to help you set up a server or something it gives you the malicious code.
Which is why people still need to learn to code. We are gonna find out real quick who the copypaster fakes are.
There was an article a week ago about how hundreds of compromised know exploits were found in GitHub.
Keep grinding your ITs hours down and outsource and contract and eventually you have 21 year olds responsible for global security with no time or resources and then hey, you’re Boeing!
I planned to selfhost my own code repos forever, this gives me one more little push, but let’s see if I move it to my server.
this persons fucks
A couple devs i know got hit by this. They are going after anyone and everyone.
But… why?
To try and take over other people’s ci/cd pipelines and inject malware into otherwise legitimate application binaries.
LastPass hack happened due to a developer logging I. On their home PV which had an outdated and vulnerable version of Plex installed. Swap outdated for “maliciously forked” and now attackers have legit code that can run for months before they use what they’ve injected to take over.
Why would somebody want to steal my login credentials‽‽
That would be silly, it would only go to accounts that aren’t theirs.
I wonder how that seven layer obfuscation works in practice. the gif in the article shows an encrypted string being decrypted (with the decryption key right there in plain sight?) and executed, and I get they try to hide it by right aligning it with ~1000 spaces, but wouldn’t that still be super obvious in a git commit diff?
Don’t use forks of repositories. Why don’t GitHub restrict forking. What will solve the problem.
Maybe just don’t blindly run code without at least checking the authenticity of the repo