A tool used to benchmark antivirus solutions against real-world malware samples.
Project description
Malbench
Malbench is a command-line tool for testing and evaluating the effectiveness of malware detection tools (such as antivirus solutions). It does this by running a set of malware samples, and checking if the malware is flagged by the detection tool we are evaluating. Malbench is built to be modular and configurable, so it can be customized to meet the specific needs of different users and environments.
About
Disclaimer
⚠ WARNING: Malbench is a tool for testing antivirus software against real-world malware samples. While it does not contain any malware on its own, it is designed to run malware samples provided by the user. Use Malbench only in secure and isolated environments that you are authorized in doing so. Do not use Malbench on a computer or network that contains sensitive information or data that you are not willing to lose and/or become compromised. By using Malbench, you acknowledge the risks and assume full responsibility for any damages that may result.
ℹ️ NOTE: As stated above: Malbench does not come with any malicious code installed. It is the user's responsibility to provide their own samples for testing purposes. This is to ensure that Malbench is used responsibly and ethically, and that users have control over the types of malware being tested.
Why Use Malbench
Malware detection tools are an essential component of any computer security strategy, but they are not foolproof. New techniques and methods are constantly being developed to evade detection. It is important to regularly test and evaluate the effectiveness of detection tools to ensure that we are keeping up with these evolving threats.
With all the different features and algorithms of modern antivirus solutions, it can be hard to find practical and objective results on what-all they defend against. Malbench can be leveraged to bulk-test known malware samples against antivirus solutions to deliver real and practical results. This is done by automatically launching multiple malware payloads on a system and seeing what samples are detected and which ones were evaded. With Malbench, users can customize their testing to meet their specific needs, selecting the malware samples they want to run, and the duration of a test.
How Antivirus is Benchmarked
Malbench launches each malware sample selected by the user and monitors its execution. If the sample completes without error, it is flagged as UNDETECTED
. If the sample is terminated with an error code, it is flagged as DETECTED
, indicating that it was killed by the antivirus solution. If the sample is still running after a designated time (defaults to 2 seconds), Malbench forcibly kills the program and flags it as UNDETECTED
, as the antivirus solution did not detect it in time.
After each sample is run, Malbench displays the detection rate. The detection rate is a percentage representing the amount of samples the antivirus successfully defended against.
$$ \text{Detection rate} = \frac{\text{Number of samples flagged as detected}}{\text{Total number of samples}} \times 100% $$
Installation
There are two ways to install Malbench, depending on whether you are a user or a developer.
For Users
To install Malbench for general use, follow these simple steps:
- Ensure that the following is installed on your system: Python 3.8.1+, and PIP (included with most Python installations).
- Install Malbench via pip by running the following command:
pip install malbench
For Developers
If you are a developer who wants to contribute to Malbench or modify the source code, you will need to follow a different set of installation steps:
- Ensure the following is installed on your system: Python 3.8.1+, PIP (included with most Python installations), Git, Poetry, and an IDE of your choice (VSCode recommended).
- Clone (or fork) the Malbench repository using Git:
git clone https://github.com/youkergav/Malbench.git
- Navigate to the project directory:
cd malbench
- Install the required dependencies using Poetry:
poetry install
Note: Please ensure the your PATH is configured for poetry - Launch the virtual environment using one of the following methods:
- For general use, the virtual environment can be opened with:
poetry shell
- Alternatively, the virtual environment can be opened in VSCode. First ensure the Python extension for VSCode is installed. Then, open the
Malbench
folder in VSCode and ensure the Python interpreter is set to use the local poetry.venv
(you may need to reload VSCode after this).
- For general use, the virtual environment can be opened with:
Usage
To use Malbench, simply run the following command inside your virtual environment:
python -m malbench /path/to/malware
Replace /path/to/malware
with your path
argument. The path
argument should be the path to the malware samples you want to test. This can be either a single file or a folder containing multiple files. Only executable files will be ran by Malbench.
By default, Malbench will show a banner when it starts, and will prompt you to confirm that you understand the risks involved before running the malware samples. You can disable the banner using the --no-banner
flag, and disable the confirmation prompt using the --no-warning
flag (as seen in the demo).
Malware samples are run one by one, and Malbench will wait for each sample to be stopped by the detection software or reach the specified 2 second timeout before moving on to the next sample. This timeout can be changed with the --timeout
flag. If a sample completes successfully, Malbench will print a red message with a [-]
prefix, indicating it wasn't stopped by the detection software. If a sample has to be forcibly terminated by the detection software, Malbench will print a green message with a [+]
prefix, indicating is was successfully detected and stopped.
A full list of arguments to use with Malbench are below. This can be displayed with --help
.
usage: malbench [-h] [-v] [-t TIMEOUT] [-nB] [-nW] [-d] path
positional arguments:
path file or folder path of malware executables
options:
-h, --help show this help message and exit
-v, --version shows malbench version number and exits
-t TIMEOUT, --timeout TIMEOUT
malware TTL before being marked as failure (2 default)
-nC, --no-color disables colored output
-nB, --no-banner hides the banner logo
-nW, --no-warning bypasses user confirmation before running
-d, --dev enables stack tracing
Contributing
Contributions to Malbench are always welcome! If you'd like to contribute, please fork the repository and create a new branch for your changes. Once you've made your changes, create a pull request and we'll review your changes. For major changes, please open an issue first to discuss what you would like to change.
Pull Request Checklist:
- New code follows similar coding styles of the existing codebase.
- New code is backed up by unit tests, with a minimum of 70% coverage.
- New code passes all linting and strict-typing rules.
- New code is well documented with docstrings and compiled in
/docs
. - Updates are made to the README when applicable.
License
This project is licensed under the GNU General Public License. See the LICENSE.md file for details.
Credits
Malbench was created by Gavin Youker.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file malbench-0.3.4.tar.gz
.
File metadata
- Download URL: malbench-0.3.4.tar.gz
- Upload date:
- Size: 477.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.4.1 CPython/3.11.2 Windows/10
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 8585d726ebb1fdb8f2f536ea26287c3214e596a62f432951612361258b8aae8f |
|
MD5 | 124eac02f4b9efd432d24cfdb7fd4d8c |
|
BLAKE2b-256 | 0292499d81315353c8f0a76758d4daf3a1a09929ea16be1d2e4eea0074e4baf6 |
File details
Details for the file malbench-0.3.4-py3-none-any.whl
.
File metadata
- Download URL: malbench-0.3.4-py3-none-any.whl
- Upload date:
- Size: 365.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.4.1 CPython/3.11.2 Windows/10
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | d954f199ce3240c061911ce63771429554bbd60626e7175532bf14340a23c73b |
|
MD5 | 0cc7774717c92507622d7cd1bc3dc574 |
|
BLAKE2b-256 | 3d2516a652182cb6f630b2997af32bf1528c04d5d09480ca6258a15bf0029102 |