Code framework and Partial experimental data of paper: OSEAF: A Robust One-Shot Environment-Aware Framework for Malware Detection Evasion
The folders are described as follows:
This folder contains the code framework for OSEAF:
\ExeBypass.pyis a OSEAF framework interface,\call_ExeBypass.pyis Call examples for interfaces。
To prevent abuse of the attack framework, we will retain some code.
-
The Anti-Sandbox function needs to be built by itself.
- The Anti-Sandbox code in
\SourceCodeFile\checkSandBox.hneeds to be added on its own
- The Anti-Sandbox code in
-
The encryption and decryption function needs to be built by itself.
- The
encryptfunction inNeedEncry.pyneeds to be added by itself, corresponding to\SourceCodeFile\decrypt.h - The
decryptfunction in\SourceCodeFile\decrypt.hneeds to be added by itself, corresponding toNeedEncry.py
- The
-
The compilation command needs to be built by itself
\SourceCodeFile\building_cl.batand\SourceCodeFile\building_link.batrepresent compile and link commands, respectively, and need to choose one compiler to add compile and link commands. (Our experiment chose Microsoft's cl compiler)
This folder records VirusTotal report results and related python scripts.
- The VT test results contain the VT report report (json format) for 140 samples, and the csv file 、
malware_adversarial_result.csvthat summarizes the results. - The VT key
apikeyand proxy URLproxyin the code need to be set by themselves
csv file containing 3500 malicious sample data sets, containing malicious sample information mal_type, sha256, size_kb, from VirusShare