Stress Testing Detection & Creativity
PyExfil was born as a PoC and kind of a playground and grew to be something a bit more. In my eyes it's still a messy PoC that needs a lot more work and testing to become
stable. The purpose of PyExfil is to set as many exfiltration, and now also communication, techniques that CAN be used by various threat actors/malware around to bypass various detection and mitigation tools and techniques. You can track changes at the official GitHub page.
Putting it simply, it's meant to be used as a testing tool rather than an actual Red Teaming tool. Although most techniques and methods should be easily ported and compiled to various operating systems, some stable some experimental, the transmission mechanism should be stable on all techniques. Clone it, deploy on a node in your organization and see which systems can catch which techniques.
pip (not necessarily the most updated):
pip install --user PyExfil
git clone https://www.github.com/ytisf/PyExfil cd PyExfil pip install --user -r requirements.txt
We recommend installing
py2exe as well so that you may cross compile various modules to a binary for easier transportation. You can do that with:
pip install py2exe
Go to the same folder where
PyExfil was cloned to and:
pip setup.py --user install
List of Techniques
- NTP Request
- DropBox LSP (Broadcast or Unicast)
- DNS over TLS
- ARP Broadcast
- GQUIC - Google Quick UDP Internet Connections (Client Hello)
- MDNS Query - Can be used as broadcast.
- AllJoyn. Name Service Protocol (IoT discovery) Version 0 ISAT.
- PacketSize. Using size of packet rather than actual data.
- UDP-Source-Port Using the source port in UDP as a transmission medium.
- CertExchange Leveraging certificate exchange function for short bursts of communication.
- DNSQ Leveraging DNS Queries for communication.
For usage per modules have a look at the USAGE file.
Although this tool was initially created as a game and later on turned to be a Red Team oriented tool, at the end of a day a major usage of
PyExfil is to test various DLP (Data Leakage Protection) systems as well as detection of intrusion. To make the latter mission simpler we have created a little module to generate fake data with a structure that matches both PII and PCI data sets. These are intended to trigger alerts while being broadcate outside of the network.
Here is how to use it:
from pyexfil.includes import CreateTestData c = CreateTestData(rows=1000, output_location="/tmp/list.csv") c.Run()
After this you can use which ever
PyExfil module you would like to try and exfiltrate the data set created. This way you can test your detection without risking exfiltrating valuable data.
We welcome it! From testing, to improving quality of code and up to entirely new methods.
For details about version look at the tags on this repository.
- Surprise on restructure (Add Another).
README.mdto two different files.
- Get a nice logo.
- Uniform calling convention for newer modules.
- Exfiltration data-set generator (PII&PCI).
Version 1.3 - Harpax:
- Adding 4 new modules.
- General fixups.
- Some old modules recoded to fit new standard.
- Full compatibility between Python2 and Python3.
Version 1.4 - ?:
- Expand physical exfiltration channels.
- Re-test servers on older modules.
- Add file manipulation class (for example, module
zipceptiondoes not fit into any existing category although currently residing under
Hopefully - Close Future
- Attempt at creating a more uniform call convention. See DOCUMENTATION.md.
- Fix that poorly written setup.py.
- Backport all old modules to new calling convention.
In the Distant Future - The Year 2000
- Add Golang/C++ support for portability.
- Extensive testing for py2exe support.
People & Companies
- Big shout out to JetBrains!!!
- Thanks to barachy and AM for ideas on protocols to use.
- Thanks to Itzik Kotler for some ideas.
- Shout out to @cac0ns3c for resolving some dependency hell.
- Thanks to @Nilesh0301 for pointing out some Python compatibility issues.
- Big thanks to @hbmartin for pointing us to pytube3 latest update and support.