ai-exploits

Exploits de IA El mundo de la IA tiene un problema de seguridad y no solo está en las entradas dadas a LLMs como ChatGPT. Basado en investigaciones realizadas por Protect AI y expertos en seguridad independientes en la Plataforma de Bug Bounty Huntr, existen ataques mucho más impactantes y...

sentrygrid

SentryGrid Plataforma de detección de amenazas impulsada por ML y monitoreo continuo construida para arquitecturas federales de confianza cero. Nativa de OSCAL, alineada con FedRAMP Moderate, auditoría encadenada por hash, ensamble de anomalías de tres modelos. Inicio rápido root@kitploit: make...

kroxylicious-pqc-filter

Filtro de cifrado de registros PQC de Kroxylicious Un complemento de filtro de Kroxylicious que proporciona cifrado transparente a nivel de registro Post-Cuántico PQC para Apache Kafka mediante encapsulación de claves ML-KEM FIPS 203 con cifrado simétrico AES-256-GCM. Los productores y consumidor...

SecML

SecML: A library for Secure and Explainable Machine Learning SecML is an open-source Python library for the security evaluation of Machine Learning ML algorithms. It comes with a set of powerful features: Wide range of supported ML algorithms. All supervised learning algorithms supported by...

vger

V'ger V'ger es una aplicación interactiva de línea de comandos para la post-explotación de instancias Jupyter autenticadas, centrada en operaciones de seguridad de IA/ML. Historias de Usuario Como Red Teamer , has encontrado credenciales de Jupyter, pero no sabes qué puedes hacer con ellas. V'ger...

H2O ImportFiles - Local File Inclusion

An attacker is able to read any file on the server hosting the H2O dashboard without any authentication. id: CVE-2023-6038 info: name: H2O ImportFiles - Local File Inclusion author: danmcinerney,byt3bl33d3r severity: high description: | An attacker is able to read any file on the server hosting t...

Awesome-MoAI-Security

Awesome Mobile On-Device AI Security SoK: Attack and Defense Landscape of Mobile On-device AI Systems Mobile on-device AI systems execute AI models locally through ML frameworks such as LiteRT/TFLite , Core ML , ExecuTorch , ONNX , and hardware-backed accelerators. This repo tracks the security...

ml-kem-key-recovery

Full key recovery from a partial Fujisaki-Okamoto comparison in wolfSSL ML-KEM CVE-2026-6330 NEON and CVE-2026-10097 AVX2. Fixed in wolfSSL 5.9.2. The full technical write-up is published as Cryptology ePrint Archive, Paper 2026/1682 local copy in paper/. See Citation below. Introduction ML-KEM...

frostbyte

FrostByte Progolue: In the past few days I've been experimenting with the AppDomain manager injection technique had a decent success with it in my previous Red Team engagements against certain EDRs. Although, this is really good for initial access vector, I wanted to release a POC which will help...

purple-team-exercise-framework

Purple Team Exercise Framework PTEF El Purple Team Exercise Framework PTEF es la metodología estándar de la industria para diseñar y ejecutar ejercicios de seguridad colaborativos que reúnen a la Inteligencia de Amenazas Cibernéticas CTI, los Red Teams y los Blue Teams. Los Purple Team Exercises...

camel-pqc-tls

TLS Post-Cuántico con Apache Camel Este repositorio contiene tres ejemplos que demuestran handshakes TLS 1.3 de Criptografía Post-Cuántica PQC con Apache Camel, utilizando el intercambio de claves híbrido X25519MLKEM768. X25519MLKEM768 combina el clásico X25519 Diffie-Hellman de curva elíptica co...

modelaudit

ModelAudit Secure your AI models before deployment. Static scanner that detects malicious code, potential backdoor indicators, and security vulnerabilities in ML model files — without ever loading or executing them. Full Documentation | Usage Examples | Supported Formats...

mlkem-native

mlkem-native mlkem-native è un'implementazione sicura, veloce e portabile in C901 di ML-KEM2. È un fork dell'implementazione di riferimento di ML-KEM3. Tutto il codice C in mlkem/src/ e mlkem/src/fips202/ è dimostrato memory-safe nessun overflow di memoria e type-safe nessun overflow di interi...

HydraDragonAV-Mobile

🐉 HydraDragonAV Mobile थ्रेट प्रोटेक्शन के साथ उन्नत एंड्रॉइड एंटीवायरस ⚠️ प्रोजेक्ट रोका गया है। इस प्रोजेक्ट को पर्याप्त रुचि नहीं मिली, इसलिए मैंने इसे कुछ समय के लिए रोक दिया है। विकास HydraDragonAntivirus पर स्थानांतरित हो गया है। डिस्कॉर्ड कम्युनिटी सर्वर: https://discord.gg/7XMCuj5mbP लोगो...

CVE-2025-10951

CVE-2025-10951 - checkmlloggertraversal.py Strumento di verifica dell'esistenza per CVE-2025-10951 / GHSA-8x9j-2p8r-7xc6 — una path traversal non autenticata in geyang/ml-logger ≤ 0.10.36, che interessa le funzioni loghandler / streamhandler in mllogger/server.py. Da utilizzare esclusivamente su...

CVE-2025-10952-ml-logger-AFR

CVE-2025-10952 — Beliebiger Dateizugriff über streamhandler in ml-logger Dieses Repository stellt ein eigenständiges Python-Skript bereit — CVE-2025-10952.py — das eine nicht authentifizierte Schwachstelle für beliebigen Dateizugriff in ml-logger's streamhandler ausnutzt, verkettet mit dem...

ML-for-RNGs

ML-for-RNGs Utilizzo del Machine Learning per prevedere le sequenze di RNG...

EUVD-2026-66477

In the Linux kernel, the following vulnerability has been resolved: gpio: ml-ioh: use rawspinlockt for the register lock iohirqtype is registered as the irqchip .irqsettype callback and takes chip-spinlock with spinlockirqsave. This callback is reached from setupirq - irqsettrigger -...

AI-Red-Teaming-Playground-Labs

KI-Red-Teaming-Playground-Labs Dieses Repository enthält die Herausforderungen Challenges für die Labs, die im Kurs „KI-Red-Teaming in der Praxis" verwendet werden. Der Kurs wurde ursprünglich auf der Black Hat USA 2024 von Dr. Amanda Minnich und Gary Lopez unterrichtet. Martin Pouliot kümmerte...

CVE-2026-80562

A flaw was found in the Linux kernel's gpio: ml-ioh component. An incorrect spinlock type spinlockt was used in interrupt handling routines where a non-sleeping lock rawspinlockt was required. This could lead to invalid lock acquisition, potentially causing system instability or a denial of...