SAEG - SAEG ：有状态自动利用生成， AE

项目说明

SAEG: Stateful Automatic Exploit Generation

<img align='right' src=' width='300px'>

SAEG is a framework uses angr as symbolic execution engine for Automatic Exploit Generation (AEG). Its purpose is to provide an efficient framework for handling multi-stage exploits that include information leakage. As a conceptual implementation, SAEG has implemented parts of stack exploitation and a prototype for heap exploitation.
We provide source code and a one-click deployment script for reproduction and reference of other studies. It also offers some benchmark test cases, and you can check the results of these test cases in the CI run results.
If you want to know more about SAEG, please refer to our article.

--------------

SAEG 是一个使用 angr 作为符号执行引擎的 AEG (Automatic Exploit Generation) 框架，旨在提供一个高效的处理包含信息泄露的多步利用的框架。作为概念验证，SAEG 实现了部分栈利用以及一个堆利用的原型，并提供了源码及一键部署环境以供复现研究，同时也提供了一些基准测试用例，你可以在 CI 中看到这些测试用例的运行结果。
如果您希望了解更多关于SAEG的信息，请查看我们的文章。

Demo video for SAEG(演示视频):

![asciicast demo video]()

Installing & Testing on Docker

It is recommended to use Ubuntu20.04 or later on docker to run the test.
Example of Docker file is based on Ubuntu22.04.
Also refer to CI file for more details.

如果您在中国大陆地区，Dockerfile 中有两行注释掉的源(apt和pip)，您可以取消注释并以加速构建。

Extra Dataset

Beyond the paper benchmarks, this repository includes an additional dataset for broader research use, which is not part of CI.
除论文基准数据外，本仓库还提供了一组额外数据，以便更广泛的研究使用；该数据集不包含在 CI 中。

SAEG-unsupported and resource-limited extra cases are kept in assets/extra_dataset/unsupported for researchers to e with other methods.
Please feel free to improve or surpass SAEG!
SAEG 暂不支持以及资源受限的额外样例保存在 assets/extra_dataset/unsupported，供研究者使用其他方法进行测试。
这里面的许多测试样例 SAEG 都不支持，请尽情玩耍，拓展 AEG 的边界！

Usage

pwn local file:

pwn remote with libc and ld specified:

Get help message:

Framework Structure

aseg.py Run framework

testset.py Test index

Dockerfile

aeg_module/

+ aeg_main.py Entry and outer state machine
+ challenge.py Initialize the binary and analyse its static information
+ binary_interactive.py Module to store label information and maintain interaction
+ mod_leak.py Attack Templates for infoleak
+ mod_exploit.py Attack Templates for stack exploit
+ mod_technique.py Monitor memory write and execution path during step() in angr
+ mod_sim_procedure.py Hook functions for angr
+ mod_sim_procedure_heap.py Hook memory functions for angr
+ utils.py Misc tools such as interaction and calculation and shellcode

SAEG was designed with the concept of scalability. If you want to simply modify/expand the framework, you may start with files starting with mod_.
>SAEG是基于易于扩展的理念设计的，如果您希望简单地修改/扩充框架，可以考虑从mod_开头的文件开始。

Additional Documentation

Citation

If you use this repository for research, please cite our paper:

Publication

Link of the draft of our paper (Neither a submitted nor a preprint):
Paper PDF

Reference of open source projects

Zeratool is a traditional symbolic execution framework for binary analysis and (stack) exploitation:

BOF_AEG is a simple stack exploitation tool based on angr: (Not a now)

Simple framework for stack exploitation and detection of memory corruption:

Framework for detection of memory corruption:

Awsome kernel AEGs (with angr):

Discussion

The performance of the CI ser by Github is fluctuating, so there may be significant de e two consecutive runs. Therefore it is more recommended to run performance tests locally.
>Github提供的CI服务器性能较为波动，即使连续两次运行也会有较大的偏差，因此更推荐在本地运行性能测试。

This framework focuses on static analysis and dynamic , so its effecti is not precise enough for all protecting mechanisms (including system-level ASLR) are disabled scenarios and utilizing dangling pointers on the stack. Currently, other solutions for work include dynamically running and interacting with GDB, but these contents are not the aim of this work (information leakage).
>该框架侧重于静态分析，动态验证，因此对于保护全关的并禁用系统ASLR的场景以及利用栈上悬垂指针的场景效果不够精确，目前其他工作的解决方案是动态调试，但这些内容并非本工作（信息泄露）的目标。

项目信息

开发语言：Python

Star 数：178

技术要点

该项目属于CTF思路方向的开源资源，代码结构清晰，适合安全研究人员学习参考。项目围绕SAEG: Stateful Automatic Exploit Generation, an AEG framework for CGC / RHG and pwnable CTF challenges.展开，提供了可运行的实现，对理解相关安全机制、工具原理与攻防技术具有较高的研究价值。

从技术实现角度看，此类项目的核心在于对底层原理的深入理解与工程化落地。开发者在研究过程中可以结合官方文档与社区讨论，梳理项目的初始化流程、核心模块职责与数据流转，逐步掌握其中的技术细节与设计思路。

使用与学习建议

1. 下载源码压缩包，解压到本地环境。

2. 根据项目类型准备对应的运行环境与依赖。

3. 按照 README 中的步骤完成配置与部署。

4. 在隔离的测试环境中运行验证，避免影响生产系统。

免责声明

本资源来源于开源社区，仅供安全研究、技术学习与合法授权范围内的攻防演练使用，请勿用于任何非法或侵害他人权益的用途。使用本项目产生的一切后果由使用者自行承担。