exploitgym - ExploitGym is a large-scale, realistic benchmark b

项目说明

ExploitGym

![Website]()
![Paper]()

ExploitGym is a large-scale, realistic benchmark built from real-world across userspace programs, Google's V8 engine, and the Linux kernel, designed to e AI agents' ability to develop exploits.

Quick start

No secret is baked into the repo: the controller generates its token salt,
flag seed, and API key on startup (or takes them from those env vars) and
both ends must use the same — see
Controller secrets.

Detailed setup steps (system dependencies, GDB, static node, agent
CLIs) live in docs/setup.md.

Documentation

Benchmark updates

The released benchmark is actively maintained. The current release is v1.0 with
869 instances. See CHANGELOG.md for the full history. The
canonical task list for the current release is data/task_ids/v1.txt.

Citation

If you use ExploitGym in your research, please cite:

License

The source code is licensed under Apache-2.0. The bundled task data
under data/tasks/ derives from external upstreams and retains their
respecti, see DATA_LICENSE.md.

项目信息

开发语言：C

Star 数：1137

技术要点

该项目属于漏洞知识库方向的开源资源，代码结构清晰，适合安全研究人员学习参考。项目围绕ExploitGym is a large-scale, realistic benchmark built from real-world vulnerabilities designed to evaluate AI agents' ability to develop exploits.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。