I have found some software on 1337 that a friend was asking for. It’s for Windows and I’m on Linux. Is test running the software inside a VM the only way to securely vet it (to see what connections it tries to make and what processes it starts)? On VirusTotal, I get one positive and 66 negatives. Preferably, I’d like to hand over the software without worrying about giving them malicious code…

  • onlinepersona@programming.devEnglish
    2·
    5 hours ago

    VirusTotal is probably the best you can do. Although, I think it belongs to Amazon or some big company now after being bought up.

  • Mordikan@kbin.earth
    8·
    12 hours ago

    One thing to remember with VirusTotal is that it should only be used as a “tool of consensus”. A lot of AV solutions on their list will false flag binaries presumably to instill fear and get people to buy their solutions over more reputable AV providers. You want the AV that works the best which many people would equate to the most flagging (as logically incomplete as that may be).

    If you are seeing 66 flags and many of those being from reputable vendors, then I would say that what they show is likely accurate. Accuracy and relevance being two different things. 100% accuracy on keygen tools doesn’t mean that it is a relevant flag to what you’re trying to do.

    You could try software solutions like LOKI (the IoC scanner) or Azul to test specific binaries, though. I believe LOKI is just doing static analysis, but Azul I know automates the sandboxing.

    • frongt@lemmy.zipEnglish
      3·
      11 hours ago

      You can also put it through sandboxes like joesandbox or cuckoo to try to trigger and detect a payload. But some malware will expect this and not execute.

      Ultimately, judge your own risk tolerance by how much you trust the source, how much you want the software, what kind of systems protection you have in place (antivirus, antimalware, permissions control, backups), and how much you can afford to have your system compromised.

      • printf("%s", name);@piefed.blahaj.zoneOPEnglish
        11·
        23 minutes ago

        Ultimately, judge your own risk tolerance by how much you trust the source, how much you want the software, what kind of systems protection you have in place (antivirus, antimalware, permissions control, backups), and how much you can afford to have your system compromised. Thanks for this! I finally decided to hand the matter over to my friend, because vetting the binaries are just… a whole project on(in?) its own. I sent them exactly what you wrote, in addition to some concrete guidance on how to setup VMs, create backups, bootable media, whathaveyou.

  • eleijeep@piefed.socialEnglish
    11·
    16 hours ago

    VirusTotal also tells you a load of stuff about file read/writes and network access etc. What did it say? And what was the positive result for?

    • printf("%s", name);@piefed.blahaj.zoneOPEnglish
      8·
      16 hours ago

      Holy crap, I had no idea! Cool!

      The one positive is a malicious.high.ml.score by Trapmine and I can’t makes heads or tails of it.

      Domains  
e592.d.akamaiedge.net
www.msn.com
www.msn.com.edgekey.net

IP  
23.50.32.0/20 (owned by Akamai Technologies, Inc.)

Execution Parents  
Crack 29.2.ZIP

      None of these sub categories under Relations give any positives.

  • Souvenir@lemmy.dbzer0.comEnglish
    5·
    15 hours ago

    Stupid question, but why can’t your friend not just run it in sandboxie with internet locked down?

    No noticeable performance drain, can’t cause too much damage.

    Or isn’t this enough?

  • atro_city@fedia.io
    103·
    16 hours ago

    The title had me very confused for a few seconds. “Are you talking about cows? Why would they need to be executed? And why would they need vets before execution?”

  • Walters, Fern@bookwyr.meEnglish
    31·
    15 hours ago

    You can also acquire the original executable, and compare the differences in a binary/hex editor. VM runs would be after you have sanitized the EXE.

    • altkey (he\him)@lemmy.dbzer0.comEnglish
      2·
      3 hours ago

      ArchiCAD uses a now-usual DRM scheme of either checking a hardware thumbstick key or seeking for a background licensing server at localhost or a set IP (always-on machine with a pool of licenses one can take one from).

      https://help.graphisoft.com/AC/25/INT/_AC25_Help/003_InstallGuide/003_InstallGuide-4.htm (is it the version OP got?)

      A cracker should either patch the requirement off completely (or drop exe, dlls for in-place replacement) or emulate a suitable key/server that is a separate app. With a lot of software I pirated with this DRM model it’s usually the second way, emulating a server with always correct responses. If that’s the case, the primary exe file can as well be original (of some distinct version?) but the crack happens to a separate background program installed with the main app.

      I think we need another look at what is actually downloaded by OP and if there is some patching process jnvolved or everything is provided already cured. Then, to see and notice what processes are running in the VM, what they try to do.

    • TheFogan@programming.devEnglish
      3·
      7 hours ago

      If you have the executable from a trusted source, then you don’t need to spend your time determining if a version you got from an unsafe location is legitimate, you’d just make a copy of the trusted original and throw out the bootleg to be safe.

    • printf("%s", name);@piefed.blahaj.zoneOPEnglish
      1·
      16 hours ago

      In theory I agree, but other than buying the original executable from a reputable vendor, I have no way of verifying the integrity and authenticity of the EXE.