Builds malware analysis Windows virtual machines so that you don’t have to.
-
Python 3.3+
-
VirtualBox or an vSphere / ESXi server
-
Install git and packer using your distribution’s packaging tool (packer is sometimes called packer-io)
-
Install vagrant from their website : https://www.vagrantup.com/downloads.html (Installing from some distributions' packaging tools have caused issues).
-
pip install
malboxes:sudo pip3 install git+https://github.com/GoSecure/malboxes.git#egg=malboxes
Note
|
Starting with Windows 10 Hyper-V is always running below the operating
system. Since VT-X needs to be operated exclusively by only one Hypervisor
this causes VirtualBox (and
malboxes) to fail. To disable Hyper-V and allow
VirtualBox to run, issue the following command in an administrative command
prompt then reboot: bcdedit /set hypervisorlaunchtype off
|
The following steps assume that you have Chocolatey installed. Otherwise, follow the manual installation procedure.
-
Install dependencies:
choco install python vagrant packer git virtualbox
-
Refresh the console
refreshenv
-
Install malboxes:
pip3 install setuptools pip3 install -U git+https://github.com/GoSecure/malboxes.git#egg=malboxes
-
Install VirtualBox, Vagrant and git
-
Install Packer, drop the packer binary in a folder in your user’s PATH like
C:\Windows\System32\
-
Install Python 3 (make sure to add Python to your environment variables)
-
Open a console (Windows-Key + cmd)
pip3 install setuptools pip3 install -U git+https://github.com/GoSecure/malboxes.git#egg=malboxes
This creates your base box that is imported in Vagrant. Afterwards you can re-use the same box several times per sample analysis.
Run:
malboxes build <template>
You can also list all supported templates with:
malboxes list
This will build a Vagrant box ready for malware investigation you can now include it in a Vagrantfile afterwards.
For example:
malboxes build win10_x64_analyst
The configuration section contains further information about what can be configured with malboxes.
malboxes spin win10_x64_analyst <name>
This will create a Vagrantfile
prepared to use for malware analysis. Move it
into a directory of your choice and issue:
vagrant up
By default the local directory will be shared in the VM on the Desktop. This
can be changed by commenting the relevant part of the Vagrantfile
.
For example:
malboxes spin win7_x86_analyst 20160519.cryptolocker.xyz
Malboxes can upload and interact with a VM on the Amazon Web serivces. To do so, follow these steps:
-
Malboxes will need a S3 bucket on AWS to upload the VM before converting it to an AMI (Amazon Machine Image). If you don’t have one, create one now.
-
Your instance also requires a security group with at least a rule allowing inbound connections for WinRM (Type: WinRM-HTTP, Protocol: TCP, Port Range: 5985, Source: host’s public IP).
-
Next, you need a
vmimport
service role configured. Follow the section named VM Import Service Role of this guide. These steps must be performed with an account that hasiam:CreateRole
andiam:PutRolePolicy
permissions. -
If the default config is used, change the hypervisor to aws and fill the mandatory options related. Otherwise, be sure to add all the options about AWS to your custom config.
-
Finally, you can follow the same steps described in the Box creation and the Per analysis instances sections to launch your instance!
Note
|
The AMI import can take a very long time (about an hour), however you can verify the status of the task by doing this. At the moment, only one AMI can be build per template. |
Install awscli using pip:
pip install awscli
Configure awscli with:
aws configure
Then run:
aws ec2 describe-import-image-tasks
To connect to an instance on the cloud using RDP, run this command at the same location of your Vagrantfile
:
vagrant rdp -- /cert-ignore
For this to work, the instance will require a security group allowing RDP inbound connections (Type: RDP, Protocol: TCP, Port Range: 3389, Source: host’s public IP).
Note
|
You can safely ignore the following error because rsync is not yet implemented: No host IP was given to the Vagrant core NFS helper. This is an internal error that should be reported as a bug.
|
Malboxes' configuration is located in a directory that follows usual operating system conventions:
-
Linux/Unix:
~/.config/malboxes/
-
Mac OS X:
~/Library/Application Support/malboxes/
-
Win 7+:
C:\Users\<username>\AppData\Local\malboxes\malboxes\
The file is named config.js
and is copied from an example file on first run.
The example configuration is documented.
Malboxes uses virtualbox as a back-end by default but since version 0.3.0 support for ESXi / vSphere has been added. Notes about the steps required for ESXi / vSphere support are available. Since everyone’s setup is a little bit different do not hesitate to open an issue if you encounter a problem or improve our documentation via a pull request.
We are exploring with the concept of profiles which are stored separately than the configuration and can be used to create files, alter the registry or install additional packages. See profile-example.js for an example configuration. This new capacity is experimental and subject to change as we experiment with it.
Currently, Malboxes does not support the automatic creation of the security groups, so you’ll have to use the AWS console to create yours. However, using the library Boto3 there should be a way to implement this.
malboxes was presented at NorthSec 2016 in a talk titled Applying DevOps Principles for Better Malware Analysis given by Olivier Bilodeau and Hugo Genesse
Code is licensed under the GPLv3+, see LICENSE
for details. Documentation
and presentation material is licensed under the Creative Commons
Attribution-ShareAlike 4.0, see docs/LICENSE
for details.
After I had the idea for an improved malware analyst workflow based on what I’ve been using for development on Linux servers (Vagrant) I quickly Googled if someone was already doing something in that regard.
I found the packer-malware repo on
github by Mark Andrew Dwyer. Malboxes was boostrapped thanks to his work which
helped me especially around the areas of Autounattend.xml
files.