Quark Engine

An Obfuscation-Neglect Android Malware Scoring System

Quark-Engine is also bundled with BlackArch. :shipit: A trust-worthy, practical tool that's ready to boost up your malware reverse engineering. https://twitter.com/quarkengine

Why Quark?

Android malware analysis engine is not a new story. Every antivirus company has their own secrets to build it. With curiosity, we develop a malware scoring system from the perspective of Taiwan Criminal Law in an easy but solid way.

We have an order theory of criminal which explains stages of committing a crime. For example, crime of murder consists of five stages, they are determined, conspiracy, preparation, start and practice. The latter the stage the more we’re sure that the crime is practiced.

According to the above principle, we developed our order theory of android malware. We develop five stages to see if the malicious activity is being practiced. They are 1. Permission requested. 2. Native API call. 3. Certain combination of native API. 4. Calling sequence of native API. 5. APIs that handle the same register. We not only define malicious activities and their stages but also develop weights and thresholds for calculating the threat level of a malware.

Malware evolved with new techniques to gain difficulties for reverse engineering. Obfuscation is one of the most commonly used techniques. In this talk, we present a Dalvik bytecode loader with the order theory of android malware to neglect certain cases of obfuscation.

Our Dalvik bytecode loader consists of functionalities such as 1. Finding cross reference and calling sequence of the native API. 2. Tracing the bytecode register. The combination of these functionalities (yes, the order theory) not only can neglect obfuscation but also match perfectly to the design of our malware scoring system.

Easy to Use and Reading Friendly Report

Quark is very easy to use and also provides flexible output formats. There are 3 types of output report: detail report, call graph, and summary report. Please see below for more details.

Detail Report

This is a how we examine a real android malware (candy corn) with one single rule (crime).

$ quark -a sample/14d9f1a92dd984d6040cc41ed06e273e.apk \
                 -r rules/ \
                 --detail

and the report will look like:

Call Graph for Every Potential Malicious Activity

We'll soon release this feature in the next version of quark!

Summary Report

Examine with rules.

quark -a sample/14d9f1a92dd984d6040cc41ed06e273e.apk \
               -r rules/ \
               --summary

QuickStart

Installation

$ git clone https://github.com/quark-engine/quark-engine.git; cd quark-engine/quark
$ pipenv install --skip-lock
$ pipenv shell

Make sure your python version is 3.7, or you could change it from Pipfile to what you have.

Check --help to see the detailed usage description.

$ quark --help

Test It Out

You may refer to the Quark Engine Document for more details of testing and development information.

Analysis Reports of Real Malware

Quark Engine will soon provide analysis reports of real malware! For your best experience of viewing the report, please use desktop web browser. We're planning to make a mobile version of report. If you really want to see the very first version of report please visit here

Also, we will soon give out our new detection rules!