My project was hacked.
When I closed my computer this random anonymous guy detected i wasn’t active and edited the code.
Its called onlinux.glitch.me.
Clarification: Well i did cd /t/t/t;, && was disabled, ; was also disabled. That meant he would be stuck in the /t/t/t directory. And he acessed the project that’s different from someone abusing the terminal.
I just saw a anonymous person logging stuff in assets/some random file name which could contain private data. He also was able to see my previous terminal history :0 and recover the file while i was inactive after a rewind. gosh he even disabled my restrictions that restrict commands but I added them back and rebackupped it.
mate, you realise the project is a literal terminal? there aren’t some nice people on here, tbh, You should’ve expected someone to do somethin like that.
Well i did cd /t/t/t;, && was disabled, ; was also disabled. That meant he would be stuck in the /t/t/t directory.
And he acessed the project that’s different from someone abusing the terminal.
@mayank1234cmd, before your posts gets flagged for name-shaming (which is not allowed on this forum), I advise you to remove the part where you accuse other users.
The only way someone can “hack” you is if they got your git url. This was most likely a staff member. If your project is warned too many times, staff may take a look inside your project. But not to jump to conclusions. This could happen for a lot of things. The probability of you being hacked is not that big. It is really small.
If this feature was seen as ‘social engineering’ it means that Glitch needs to remove under the rules set by their hosting provider, AWS. Could you tell us exactly how the feature worked? Did it contain anything that could be against Glitch or AWS ToS?
<?php
// Lemme think here for a bit, defending an online terminal from the directory commands is atrocious work
// idea: idea! convert command to array and check if array[1] is == cd, rm, or array contains > or >>
//hmm, could be doable, but at the same time, maybe not s=_GET[‘command’];
$s=explode(" ", $s);
echo $s;
I wasn’t able to really get much from the current state of the project, but if the above mentions of accepting user input and running it as a terminal command is true, then that’s a serious security issue that you can’t really work around.
Escaping input or detecting certain malicious commands is not enough, as people have already mentioned. I’d advise against making any app that takes user input and runs it as a terminal command within your app’s container.
mate, I’d suggest make it clonable, and if the project url is not equal to “onlinux.glitch.me” actually show the terminal, that way the core project can’t be deletable