Pages

Saturday, September 3, 2016

Cookie Cruncher Update, Timelines, Chrome Parser and more

I just wanted to pass on that I had a chance to update my Google Analytic Cookie Cruncher to support Firefox up to version 48. I can't believe it's been two years since I've updated the code!

I know I've said it before, but if you need me to update a tool to support a newer version of "X", please let me know - I'm happy to do so :) With everything else on my plate, I don't always have time to test each new browser for compatibility issues. Thanks to Heather Mahalik for reaching out to me with a student request to get it updated - sometimes I need that extra motivation.

I also updated my script that parses Google Analytics from Safari binary cookies. Mike O'Daniel reached out to me when the script crashed on him. Although he was unable to share the data due to privacy reasons, with a little back and forth trouble shooting we were able to determine what the issue was. He was parsing cookies from an iPad which contained URL encoded strings. None of my test data contained cookies formatted in this way and I did not have access to an iPad. Once the issue was fixed in the script he was off an running. Thanks to Mike for reaching out  to me to let me know that there were issues, and taking the time to help trouble shoot it since I was not able to replicate the issue.

I also wanted to push out a simple little parser for Chrome Internet History and Downloads. I recently spoke at the HTCIA conference  about mini-timelines (and even micro timelines). While this concept is nothing new, I have found this process to be invaluable during the cases I work. Harlan has blogged many times about the process and advantages of it, so I won't go into detail here. For the lab I taught, I just needed to output some basic Chrome Internet History into TLN format so I wrote a Chrome parser in python.

Now this tool does not show every single thing that is available in the Chrome History. I just stuck to the basic information: Visit time, URL, Hit Count etc. Sometimes too much information can cloud the timeline, making it difficult to pick out patterns of activity, or create so much noise the next lead gets lost in all the output.

I like the data in my timeline to be concise and clear. It reminds me a little of keyword searching. If the term is vague, you may be casting a wider net, but relative results could get buried in a million hits. It's going to take a lot of sifting to find that golden nugget. However, if you use a carefully crafted keyword, you can focus in on what it is you are looking for. Timelines are the same. Carefully picking the artifacts you want to add in to the timeline can help you hone in on relevant data quickly.

The other thing I wanted to discuss was Volatility plugins. I recently had the chance to run through a demo at a Python Meetup group on what Memory forensics is, and how Volatility can be used to analyze memory. As part of this, I "wrote" my first volatility plugin. Now, I say "wrote" because it was really just modifying a couple of lines in someone else's code to do something a little different.

Volatility has provided a nice interface to grab various keys from the registry. In fact, it reminds me of the way plugins are handled in RegRipper. If there is a key that you want that is not currently supported, look for a plugin that is similar and see if you can tweak it. It's a great way to start out, and as you tweak more and add a little bit here and there, you being to understand how things work.

I just started with something simple - pulling the computer name. This is just one key, with no binary data to convert:

HKLM\CurrentControlSet001\Control\ComputerName\ComputerName



I found another Volatility plugin that pulls a key from the system hive, shutdown.py - changed a few lines of code, and et voila! My first plugin. Ok - nothing earth shattering or difficult, but it's the first step in understanding how things work. That's often the way that I write many of my scripts - break it down into pieces, find code examples, and put it all together. Pretty soon I actually remember some of it, and my skill set advances.

The original code was written by Jamie Levy (@gleeda), and pulls the shutdown time from the registry. Below is a example of what I did. I just commented out what I didn't need, and modified what I did need.

While it may not be complex, it gets the job done and I learned something new in the process.

No comments:

Post a Comment