Any Powershell gurus here?

[airchie]

Hi guys,

As part of my SOC Analyst role I often come across malicious PS scripts.
I often have to try to deobfuscate them before I can try to work out what they're doing which is one of the questions I want to ask.
Once I have deobfuscated them enough to get a feel for what they're doing, they almost always create a new browsing object to download a malicious payload.

So I have a couple of questions I'd like to pose tot he group:
1. Is there any easy way to get scripts deobfuscated?
The machine obviously needs to deobfuscate them to know which cmdlets to invoke with which parameters etc.

2. Is there any way to "disable" or monitor specific cmdlets?
Particularly the ones that create new web objects?

Thanks in advance.

 

[PHolder]

You should have a visit over at SANS institute. They seem to post quite frequently about deobfuscation. https://isc.sans.edu/

 

[Telmeanius]

Hi Archie.
Were I in your shoes I would start with a execution-policy change.
Powershell 101.
Know who has rights to run a script.
PS C:\Windows> Get-ExecutionPolicy -List

Scope ExecutionPolicy
----- ---------------
MachinePolicy Undefined
UserPolicy Undefined
Process Undefined
CurrentUser Undefined
LocalMachine Bypass

Now set that policy by scope of user.
Set-ExecutionPolicy -ExecutionPolicy Restricted -Scope CurrentUser

Needless to say the this is a big hammer and its going to make sysadmin jobs tougher.
Stop it all,,, where you can,, and at least the target get much smaller.

Monitoring command-lets.
Try adding Start-Transcript to the default profile.
Assuming "they" dont think about Stopping it, This should capture all actions performed in a PS instance.

Start-Transcript (Microsoft.PowerShell.Host) - PowerShell

The Start-Transcript cmdlet creates a record of all or part of a PowerShell session to a text file. The transcript includes all command that the user types and all output that appears on the console. By default, Start-Transcript stores the transcript in the following location using the default...

docs.microsoft.com

Tel

 

[nddrifter]

You cannot get much more authoritative on this subject than Lee Holmes. See https://github.com/danielbohannon/Revoke-Obfuscation for the script that pairs with Daniel's Invoke-Obfuscation script. Based on how you are using it, you will likely also be interested in a whitepaper about why he wrote it and how you can use it at https://www.blackhat.com/docs/us-17...on-Detection-And Evasion-Using-Science-wp.pdf. Their presentation would be nice to understand the flexibility of the parser.

 

[SaltyOldGeek]

Agree, you could change the execution policy. You could also try to setup a way to scrape and report on the Powershell event logs.

 

[rfrazier]

I cannot answer the OP original question. But, as a simple one person end user, I prefer to disable PowerShell completely although I haven't actually done that in Win 7. But, my feeling is that I just don't need the functionality nor the attack surface. Same as Java years ago. Of course, this may not actually be possible. If it's possible, in corporate settings, this may not be practical.

May your bits be stable and your interfaces be fast. Ron

 

[SaltyOldGeek]

A lot of rmm(remote management) services and applications use Powershell. You can, however, set ACL's(still learning how myself) though in most enterprise applications you run a Powershell script via an rmm agent and disable remote PowerShell altogether. Dato and Pulseway both use Powershell scripts but run them locally via the agent with secure transfer of the script.