GnomeMan4201/BANANA_TREE

GitHub: GnomeMan4201/BANANA_TREE

BANANA_TREE 是一个对抗性训练生态系统,通过模拟真实攻击来提升安全防御能力。

Stars: 1 | Forks: 0

``` ██████╗ █████╗ ███╗ ██╗ █████╗ ███╗ ██╗ █████╗ ████████╗██████╗ ███████╗███████╗ ██╔══██╗██╔══██╗████╗ ██║██╔══██╗████╗ ██║██╔══██╗ ╚══██╔══╝██╔══██╗██╔════╝██╔════╝ ██████╔╝███████║██╔██╗ ██║███████║██╔██╗ ██║███████║ ██║ ██████╔╝█████╗ █████╗ ██╔══██╗██╔══██║██║╚██╗██║██╔══██║██║╚██╗██║██╔══██║ ██║ ██╔══██╗██╔══╝ ██╔══╝ ██████╔╝██║ ██║██║ ╚████║██║ ██║██║ ╚████║██║ ██║ ██║ ██║ ██║███████╗███████╗ ╚═════╝ ╚═╝ ╚═╝╚═╝ ╚═══╝╚═╝ ╚═╝╚═╝ ╚═══╝╚═╝ ╚═╝ ╚═╝ ╚═╝ ╚═╝╚══════╝╚══════╝ ```
[![生态系统](https://img.shields.io/badge/BANANA__TREE-ADVERSARIAL%20ECOSYSTEM-8B1A1A?style=for-the-badge&labelColor=0A0A08)](https://github.com/GnomeMan4201) [![集体](https://img.shields.io/badge/LANimals-COLLECTIVE-5C1010?style=for-the-badge&labelColor=0A0A08)](https://github.com/GnomeMan4201) [![理念](https://img.shields.io/badge/GOOD%20DEFENSE-IS%20GOOD%20OFFENSE-3DAA3D?style=for-the-badge&labelColor=0A0A08)](https://github.com/GnomeMan4201)
## 查看 /etc/banana_tree/philosophy.txt 的内容 ``` BANANA_TREE exists to teach the adversarial mindset. not designed....but evolved out of NECESSITY every tool in this ecosystem was built to close a real gap in a real problem at a real moment. laid side by side they form a complete end-to-end adversarial training environment. that was not the plan. but became part of it.. ``` ## 列出 ./ecosystem 的详细信息 ``` BANANA_TREE/ │ ├── roots/ # observe — know your environment │ ├── LANimals │ ├── OpenSight │ └── devto-analytics-pro │ ├── trunk/ # simulate — model the threat │ ├── Lune │ ├── Decoy-Hunter │ └── bad_BANANA │ ├── branches/ # execute — apply the pressure │ ├── zer0DAYSlater │ ├── OWN │ └── Blackglass_Suite │ └── fruit/ # adapt — close the loop ├── reflexive-identity ├── devto-bot-audit └── badBANANA zine // Issue 01 ``` ## 列出 ./roots/ 的详细信息 ### 🌿 LANimals 实时局域网映射、异常检测、行为风险评分、 蜜罐陷阱引擎、力导向图可视化。 这是基础——无法被看见的东西就无法被防御。 ``` role: observation layer — maps the environment in real time feeds: threat surface data into simulation and execution layers stack: Python · FastAPI · SQLite · D3.js · SQLAlchemy ``` [![LANimals](https://img.shields.io/badge/→%20LANimals-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/LANimals) ### 🌿 OpenSight 实体提取、类型化知识图谱、社区检测、 LLM 摘要、调查包系统。 将原始文档转化为结构化情报。 ``` role: intelligence layer — extracts signal from unstructured data feeds: entity and relationship data into analysis and adaptation stack: Python · FastAPI · SQLAlchemy · Canvas API · SQLite ``` [![OpenSight](https://img.shields.io/badge/→%20OpenSight-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/OpenSight) ### 🌿 devto-analytics-pro 追踪开发者平台的参与度、增长信号和行为模式。 针对公开画像的表层OSINT。 ``` role: signal layer — monitors public-facing behavioral patterns stack: Python · FastAPI · SQLite ``` [![devto-analytics](https://img.shields.io/badge/→%20devto--analytics--pro-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/devto-analytics-pro) ## 列出 ./trunk/ 的详细信息 ### 🌿 Lune 64模块的谍报技术库。由LLM驱动的突变引擎。 预构建的操作链。加密的C2(AES-128-CBC + HMAC-SHA256)。 统一的身份系统。具有加密信标/报告周期的加固代理。 ``` role: simulation layer — models adversarial behavior end-to-end note: authorized red team ops and controlled lab simulation only stack: Python · MIT licensed · CI passing ``` [![Lune](https://img.shields.io/badge/→%20Lune-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/Lune) ### 🌿 Decoy-Hunter 主动欺骗部署与检测工具。 布置陷阱。观察其触发。学习是什么触发了它。 ``` role: deception layer — creates false surfaces, detects contact feeds: behavioral data back into observation layer stack: Python ``` [![Decoy-Hunter](https://img.shields.io/badge/→%20Decoy--Hunter-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/Decoy-Hunter) ### 🌿 bad_BANANA 适用于 Android (Termux) + Debian 的无需root权限的攻击性工具包。 为在传统实验室环境之外工作的操作员而构建。 ``` role: mobile simulation layer — extends the environment off-network stack: Python · Shell ``` [![bad_BANANA](https://img.shields.io/badge/→%20bad__BANANA-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/bad_BANANA) ## 列出 ./branches/ 的详细信息 ### 🌿 zer0DAYSlater AI驱动的有效载荷突变、自适应持久化、 实时漏洞利用技术。 ``` role: execution layer — applies offensive pressure to test defense note: authorized simulation environments only stack: Python ``` [![zer0DAYSlater](https://img.shields.io/badge/→%20zer0DAYSlater-0A0A08?style=flat-square&labelColor=5C1010)](https://github.com/GnomeMan4201/zer0DAYSlater) ### 🌿 OWN ``` role: execution layer — payload staging and persistence testing note: authorized simulation environments only stack: Python ``` [![OWN](https://img.shields.io/badge/→%20OWN-0A0A08?style=flat-square&labelColor=5C1010)](https://github.com/GnomeMan4201/OWN) ### 🌿 Blackglass Suite ``` role: execution layer — operational support and environment setup stack: Shell · Bash ``` [![Blackglass](https://img.shields.io/badge/→%20Blackglass__Suite-0A0A08?style=flat-square&labelColor=5C1010)](https://github.com/GnomeMan4201/Blackglass_Suite) ## 列出 ./fruit/ 的详细信息 ### 🌿 reflexive-identity 针对自主代理的持续信任评分。 当系统质疑自身时——那就是适应层。 ``` role: adaptation layer — closes the loop, updates the model stack: Python ``` [![reflexive-identity](https://img.shields.io/badge/→%20reflexive--identity-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/reflexive-identity) ### 🌿 badBANANA // Issue 01 — 生命系统 从运行此生态系统中学到的一切—— 浓缩成一份安全研究出版物。 树之果实。 ``` topics: adversarial AI · prompt-space security · network deception living off the land · red team ops release: DEF CON format: monospace · cream / black / maroon ``` [![阅读 Issue 01](https://img.shields.io/badge/→%20badBANANA%20Issue%2001-0A0A08?style=flat-square&labelColor=8B1A1A)](https://github.com/GnomeMan4201/bad_BANANA/releases/tag/issue-01) ## 查看 ./origin.txt 的内容 ``` BANANA_TREE was not planned. each tool was built alone, in the moment, to solve a problem that existed right then. LANimals because the network was invisible. OpenSight because the documents were unreadable. Lune because simulation required a real engine. Decoy-Hunter because traps needed to be set. zer0DAYSlater because defense needed real pressure. reflexive-identity because the agent needed to doubt itself. laid side by side: a complete adversarial training environment. that is what emergence looks like. that is the adversarial mindset in practice — built the tools you needed, discovered the system you had. ```
``` good defense takes good offense ``` *© badBANANA collective // LANimals // gnomeman4201*
标签:Cutter, TGT, 人工智能安全, 反取证, 合规性, 安全评估, 密钥泄露防护, 对抗训练, 应用安全, 攻击模拟, 攻防演练, 数据展示, 机器学习安全, 生态系统, 红队, 网络安全, 逆向工具, 隐私保护, 集体协作, 驱动签名利用