entropystrip

___________       __                               _________ __         .__        
\_   _____/ _____/  |________  ____ ______ ___.__./   _____//  |________|__|_____  
 |    __)_ /    \   __\_  __ \/  _ \\____ <   |  |\_____  \\   __\_  __ \  \____ \ 
 |        \   |  \  |  |  | \(  <_> )  |_> >___  |/        \|  |  |  | \/  |  |_> >
/_______  /___|  /__|  |__|   \____/|   __// ____/_______  /|__|  |__|  |__|   __/ 
        \/     \/                   |__|   \/            \/                |__|    

What it does

Most modern malware hides behind packers (like UPX, VMProtect, or Themida) or stashes encrypted payloads in PE resource directories or trailing file overlays. Rather than relying on static byte signatures, entropystrip evaluates:

1. Section Entropy: Calculates block-level Shannon entropy across every PE section to locate encrypted or compressed buffers ($H > 7.20$).
2. Memory Permissions: Identifies sections marked simultaneously as Writable and Executable (IMAGE_SCN_MEM_WRITE and IMAGE_SCN_MEM_EXECUTE), which indicates self-modifying code or runtime unpackers.
3. PE Overlays: Measures trailing bytes appended past the end of the last mapped section.
4. Import Table Size & Composition: Flags binaries with unusually small IATs (e.g. only importing LoadLibrary / GetProcAddress) or sensitive process injection primitives (VirtualAlloc, WriteProcessMemory, QueueUserAPC).
5. .NET Classification: Inspects the COM descriptor directory to separate managed CLR assemblies from native binaries, avoiding false positives on standard C# applications.
6. Payload Carving: Extracts identified high-entropy sections or overlays straight to raw binary files on disk for further analysis.
7. Directory Triage: Walks directories or drives and ranks all found binaries by threat score.

Installation

Requirements: Python 3.8+ on Windows, Linux, or macOS.

pip install entropystrip

Or install directly from source:

git clone https://github.com/dwexi/entropystrip.git
cd entropystrip
pip install -e .

Usage

usage: main.py [-h] [-r] [--dump] [--deep] [--html HTML] [--json JSON] target

EntropyStrip: PE De-obfuscator & Triage Engine

positional arguments:
  target         Path to Windows executable (.exe / .dll) or directory to scan

options:
  -h, --help     show this help message and exit
  -r, --recursive
                 recursively scan subdirectories if target is a folder
  --dump         carve and write high-entropy payloads to extracted_payloads/
  --deep         deep forensic triage: MITRE ATT&CK mapping, entry point disassembly, hashes
  --html HTML    write minimalist HTML report to specified path
  --json JSON    write machine-readable JSON telemetry to specified path

Examples

Analyze a single binary:

python main.py target.exe

Deep forensic triage with MITRE ATT&CK mapping & disassembly:

python main.py suspicious.exe --deep

Analyze and extract hidden payload buffers:

python main.py suspicious.exe --dump

Carved payloads are written to extracted_payloads/<target>_<section>.bin.

Export a standalone HTML report:

python main.py sample.exe --html report.html

Triage an entire directory of unknown files:

python main.py "C:\IncidentTriage" -r

Entropy Scale Reference

Shannon entropy measures information density from 0.00 to 8.00 bits per byte:

$$H(X) = -\sum_{i=0}^{255} P(x_i) \log_2 P(x_i)$$

License

MIT License. See LICENSE for details.