Login
This repo contains the code for our USENIX Security '23 paper "ARGUS: A Framework for Staged Static Taint Analysis of GitHub Workflows and Actions". Argus is a comprehensive security analysis tool specifically designed for GitHub Actions. Built with an aim to enhance the security of CI/CD workflows, Argus utilizes taint-tracking techniques and an impact classifier to detect potential vulnerabilities in GitHub Action workflows.
Visit our website - secureci.org for more information.
Taint-Tracking: Argus uses sophisticated algorithms to track the flow of potentially untrusted data from specific sources to security-critical sinks within GitHub Actions workflows. This enables the identification of vulnerabilities that could lead to code injection attacks.
Impact Classifier: Argus classifies identified vulnerabilities into High, Medium, and Low severity classes, providing a clearer understanding of the potential impact of each identified vulnerability. This is crucial in prioritizing mitigation efforts.
This Python script provides a command line interface for interacting with GitHub repositories and GitHub actions.
FreshRSS python argus.py --mode [mode] --url [url] [--output-folder path_to_output] [--config path_to_config] [--verbose] [--branch branch_name] [--commit commit_hash] [--tag tag_name] [--action-path path_to_action] [--workflow-path path_to_workflow]--mode: The mode of operation. Choose either 'repo' or 'action'. This parameter is required.--url: The GitHub URL. Use USERNAME:TOKEN@URL for private repos. This parameter is required.--output-folder: The output folder. The default value is '/tmp'. This parameter is optional.--config: The config file. This parameter is optional.--verbose: Verbose mode. If this option is provided, the logging level is set to DEBUG. Otherwise, it is set to INFO. This parameter is optional.--branch: The branch name. You must provide exactly one of: --branch, --commit, --tag. This parameter is optional.--commit: The commit hash. You must provide exactly one of: --branch, --commit, --tag. This parameter is optional.--tag: The tag. You must provide exactly one of: --branch, --commit, --tag. This parameter is optional.--action-path: The (relative) path to the action. You cannot provide --action-path in repo mode. This parameter is optional.--workflow-path: The (relative) path to the workflow. You cannot provide --workflow-path in action mode. This parameter is optional.To use this script to interact with a GitHub repo, you might run a command like the following:
python argus.py --mode repo --url https://github.com/username/repo.git --branch masterThis would run the script in repo mode on the master branch of the specified repository.
Argus can be run inside a docker container. To do so, follow the steps:
results folderYou can view SARIF results either through an online viewer or with a Visual Studio Code (VSCode) extension.
Online Viewer: The SARIF Web Viewer is an online tool that allows you to visualize SARIF files. You can upload your SARIF file (argus_report.sarif) directly to the website to view the results.
VSCode Extension: If you prefer to use VSCode, you can install the SARIF Viewer extension. After installing the extension, you can open your SARIF file (argus_report.sarif) in VSCode. The results will appear in the SARIF Explorer pane, which provides a detailed and navigable view of the results.
Remember to handle the SARIF file with care, especially if it contains sensitive information from your codebase.
If there is an issue with needing the Github authorization for running, you can provide username:TOKEN in the GITHUB_CREDS environment variable. This will be used for all the requests made to Github. Note, we do not store this information anywhere, neither create any thing in the Github account - we only use this for cloning the repositories.
Argus is an open-source project, and we welcome contributions from the community. Whether it's reporting a bug, suggesting a feature, or writing code, your contributions are always appreciated!
If you use Argus in your research, please cite our paper:
@inproceedings{muralee2023Argus,
title={ARGUS: A Framework for Staged Static Taint Analysis of GitHub Workflows and Actions},
author={S. Muralee, I. Koishybayev, A. Nahapetyan, G. Tystahl, B. Reaves, A. Bianchi, W. Enck,
A. Kapravelos, A. Machiry},
booktitle={32st USENIX Security Symposium (USENIX Security 23)},
year={2023},
}
Tai-e (Chinese: ε€ͺιΏ; pronunciation: [ΛtaΙͺΙ:]) is a new static analysis framework for Java (please see our technical report for details), which features arguably the "best" designs from both the novel ones we proposed and those of classic frameworks such as Soot, WALA, Doop, and SpotBugs. Tai-e is easy-to-learn, easy-to-use, efficient, and highly extensible, allowing you to easily develop new analyses on top of it.
Currently, Tai-e provides the following major analysis components (and more analyses are on the way):
clone() detectorTai-e is developed in Java, and it can run on major operating systems including Windows, Linux, and macOS.
The simplest way is to download it from GitHub Releases.
Alternatively, you might build the latest Tai-e yourself from the source code. This can be simply done via Gradle (be sure that Java 17 (or higher version) is available on your system). You just need to run command gradlew fatJar, and then the runnable jar will be generated in tai-e/build/, which includes Tai-e and all its dependencies.
We are hosting the documentation of Tai-e on the GitHub wiki, where you could find more information about Tai-e such as Setup in IntelliJ IDEA , Command-Line Options , and Development of New Analysis .
In addition, we have developed an educational version of Tai-e where eight programming assignments are carefully designed for systematically training learners to implement various static analysis techniques to analyze real Java programs. The educational version shares a large amount of code with Tai-e, thus doing the assignments would be a good way to get familiar with Tai-e.
Appshark is a static taint analysis platform to scan vulnerabilities in an Android app.
Appshark requires a specific version of JDK -- JDK 11. After testing, it does not work on other LTS versions, JDK 8 and JDK 16, due to the dependency compatibility issue.
We assume that you are working in the root directory of the project repo. You can build the whole project with the gradle tool.
$ ./gradlew build -x test After executing the above command, you will see an artifact file AppShark-0.1.1-all.jar in the directory build/libs.
Like the previous step, we assume that you are still in the root folder of the project. You can run the tool with
$ java -jar build/libs/AppShark-0.1.1-all.jar config/config.json5The config.json5 has the following configuration contents.
{
"apkPath": "/Users/apks/app1.apk",
"out": "out",
"rules": "unZipSlip.json",
"maxPointerAnalyzeTime": 600
} Each JSON field is explained below.
If you provide a configuration JSON file which sets the output path as out in the project root directory, you will find the result file out/results.json after running the analysis.
Below is an example of the results.json.
{
"AppInfo": {
"AppName": "test",
"PackageName": "net.bytedance.security.app",
"min_sdk": 17,
"target_sdk": 28,
"versionCode": 1000,
"versionName": "1.0.0"
},
"SecurityInfo": {
"FileRisk": {
"unZipSlip": {
"category": "FileRisk",
"detail": "",
"model": "2",
"name": "unZipSlip",
"possibility": "4",
"vulners": [
{
"details": {
"position": "<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolderFix1(java.lang.String,java.lang.String)>",
"Sink": "<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolderFix1(java.lang.String,java.lang.String)>->$r31",
"entryMethod": "<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void f()>",
"Source": "<net.byte dance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolderFix1(java.lang.String,java.lang.String)>->$r3",
"url": "/Volumes/dev/zijie/appshark-opensource/out/vuln/1-unZipSlip.html",
"target": [
"<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolderFix1(java.lang.String,java.lang.String)>->$r3",
"pf{obj{<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolderFix1(java.lang.String,java.lang.String)>:35=>java.lang.StringBuilder}(unknown)->@data}",
"<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolderFix1(java.lang.String,java.lang.String)>->$r11",
"<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolderFix1(java.lang.String,java.lang.String)>->$r31"
]
},
"hash": "ec57a2a3190677ffe78a0c8aaf58ba5aee4d 2247",
"possibility": "4"
},
{
"details": {
"position": "<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolder(java.lang.String,java.lang.String)>",
"Sink": "<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolder(java.lang.String,java.lang.String)>->$r34",
"entryMethod": "<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void f()>",
"Source": "<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolder(java.lang.String,java.lang.String)>->$r3",
"url": "/Volumes/dev/zijie/appshark-opensource/out/vuln/2-unZipSlip.html",
"target": [
"<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolder(java.lang.String,java.lang.String)>->$r3",
"pf{obj{<net.bytedance.security.a pp.pathfinder.testdata.ZipSlip: void UnZipFolder(java.lang.String,java.lang.String)>:33=>java.lang.StringBuilder}(unknown)->@data}",
"<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolder(java.lang.String,java.lang.String)>->$r14",
"<net.bytedance.security.app.pathfinder.testdata.ZipSlip: void UnZipFolder(java.lang.String,java.lang.String)>->$r34"
]
},
"hash": "26c6d6ee704c59949cfef78350a1d9aef04c29ad",
"possibility": "4"
}
],
"wiki": "",
"deobfApk": "/Volumes/dev/zijie/appshark-opensource/app.apk"
}
}
},
"DeepLinkInfo": {
},
"HTTP_API": [
],
"JsBridgeInfo": [
],
"BasicInfo": {
"ComponentsInfo": {
},
"JSNativeInterface": [
]
},
"UsePermissions": [
],
"DefinePermis sions": {
},
"Profile": "/Volumes/dev/zijie/appshark-opensource/out/vuln/3-profiler.json"
}
Aura is a static analysis framework developed as a response to the ever-increasing threat of malicious packages and vulnerable code published on PyPI.
Project goals:
Feature list:
Didn't find what you are looking for? Aura's architecture is based on a robust plugin system, where you can customize almost anything, ranging from a set of data analyzers, transport protocols to custom out formats.
# Via pip:
pip install aura-security[full]
# or build from source/git
poetry install --no-dev -E full
Or just use a prebuild docker image sourcecodeai/aura:dev
docker run -ti --rm sourcecodeai/aura:dev scan pypi://requests -v
Aura uses a so-called URIs to identify the protocol and location to scan, if no protocol is used, the scan argument is treated as a path to the file or directory on a local system.
Diff packages:
docker run -ti --rm sourcecodeai/aura:dev diff pypi://requests pypi://requests2
Find most popular typosquatted packages (you need to call aura update to download the dataset first):
aura find-typosquatting --max-distance 2 --limit 10
While there are other tools with functionality that overlaps with Aura such as Bandit, dlint, semgrep etc. the focus of these alternatives is different which impacts the functionality and how they are being used. These alternatives are mainly intended to be used in a similar way to linters, integrated into IDEs, frequently run during the development which makes it important to minimize false positives and reporting with clear actionable explanations in ideal cases.
Aura on the other hand reports on ** behavior of the code**, anomalies, and vulnerabilities with as much information as possible at the cost of false positive. There are a lot of things reported by aura that are not necessarily actionable by a user but they tell you a lot about the behavior of the code such as doing network communication, accessing sensitive files, or using mechanisms associated with obfuscation indicating a possible malicious code. By collecting this kind of data and aggregating it together, Aura can be compared in functionality to other security systems such as antivirus, IDS, or firewalls that are essentially doing the same analysis but on a different kind of data (network communication, running processes, etc).
Here is a quick overview of differences between Aura and other similar linters and SAST tools:
# nosec that will suppress the alert at that positionAura framework is licensed under the GPL-3.0. Datasets produced from global scans using Aura are released under the CC BY-NC 4.0 license. Use the following citation when using Aura or data produced by Aura in research:
@misc{Carnogursky2019thesis,
AUTHOR = "CARNOGURSKY, Martin",
TITLE = "Attacks on package managers [online]",
YEAR = "2019 [cit. 2020-11-02]",
TYPE = "Bachelor Thesis",
SCHOOL = "Masaryk University, Faculty of Informatics, Brno",
SUPERVISOR = "Vit Bukac",
URL = "Available at WWW <https://is.muni.cz/th/y41ft/>",
}
