Pages

Sunday, June 14, 2020

Plugin Spotlight - printer_settings, featureusage

Given the number of RegRipper plugins that are part of the distro, I thought it would be a good idea every now and then to spotlight a plugin or two, and share what led to the plugin being created, and discuss how it can be used as part of analysis.

printer_settings.pl
This plugin is a result of what I read about Project TajMahal. If you scroll down in Appendix II, to modules 65 and 66, you'll see the following statement:

Steals printed documents from spooler queue.

This is done by enabling the “KeepPrintedJobs” attribute for each (or just one) configured printer stored in Windows Registry. What this means is that print jobs will not be deleted once they're complete; as such, this serves as an interesting means of data collection, specifically, data from information repositories.

I thought that was interesting and tried setting the attribute via the UI, and then writing and testing a plugin to detect the attribute setting.  The result is the plugin.

So, how would you use this during an engagement?  A positive finding from the plugin would be a pivot point into deeper analysis; for example, if the attribute is set, what is the LastWrite time of the key (or keys) in question?  Does this time stamp then prove to be a useful pivot point within the greater context of an overall system timeline?  If you have an image of the system, what is the content of the spooler?

featureusage.pl
CrowdStrike recently posted an article on the various values and subkeys beneath the FeatureUsage key, so I'm not sure what I could add to that.

In short, the FeatureUsage artifacts reportedly serve as evidence of program execution, on Windows 10 version 1903 and higher.  The CrowdStrike blog post provides some very good information regarding the subkey contents; what really stood out for me is how the contents provide insight into humanness within the Windows Registry, as well as provide information that analysts can look to in the face of counter-forensics.


No comments:

Post a Comment