binary-ninja-headless-mcp

Binary Ninja Headless MCP A headless Binary Ninja server that speaks MCP Model Context Protocol, giving AI agents full access to deep reverse-engineering workflows — disassembly, IL, patching, types, xrefs, and more — without a GUI. Designed to run in the same Docker container as the agent runtim...

CVE-2026-116690

The MLflow AI engineering platform (versions 3.10.0 through 3.17.0) is vulnerable to an arbitrary file read due to improper path validation in the _create_model_version() function. The prompt branch incorrectly accepts schemeless relative sources because validation is only triggered when a URL sc...

VulnLLM-R

VulnLLM-R: Specialized Reasoning LLM for Vulnerability Detection Paper: arXiv:2512.07533 Code & Data: GitHub Demo: Web demo Model: 7B Model Environment and dataset 🛠️ Create environment Install Git LFS and clone the repository LFS files are fetched automatically during clone git lfs install...

ExploitTensorflowCVE-2021-37678

CVE-2021-37678 Exploit To exploit the vulnerability, it is necessary to start a Docker container with an environment prepared to run Tensorflow with version 2.2.3 which still has the vulnerability. It consists of: An AMD 64-bit architecture, CPU or GPU that can execute AVX instructions Python...

CVE-2026-116686

The Model Context Protocol Python SDK contains a vulnerability in versions 1.10.0 through 1.30.0 and 2.2.0 . The ClientSession, ClientSessionGroup, and Client tool calls use the default resolver in jsonschema to validate server-declared outputSchema documents. This allows a server to provide an u...

CVE-2026-116685

The Model Context Protocol Python SDK is vulnerable in versions 1.8.0 up to 1.30.0 and 2.2.0 due to how its HTTP client transports handle redirects. Specifically, the streamable_http_client, streamablehttp_client, sse_client, ClientSessionGroup, OAuth, and create_mcp_http_client paths may follow ...

OllamaHound

OllamaHound Modular toolkit for discovering exposed Ollama servers, fingerprinting versions, and validating model access. OllamaHound is a modular, RedTeam oriented toolkit for discovering publicly exposed Ollama servers , fingerprinting versions, classifying CVE exposure , validating whether tex...

JoySafety

JoySafety — Large Model Security Framework 📋 Table of Contents Project Introduction ✨ Features 🚀 Quick Start 📖 User Guide 🏆 Best Practices 🏗️ Architecture Design 🛠️ Development 📄 License 📅 Roadmap Star History new release ✅ Prompt injection detection model fully upgraded : Training system...

CVE-2026-116684

The Model Context Protocol Python SDK versions 1.10.0 through 1.30.0 and 2.2.0 are vulnerable to a resource-confusion flaw in the server-side bearer authentication. When configured via AuthSettings with token_verifier, the TokenVerifier fails to independently validate the aud audience , causing t...

heaphopper

heaphopper HeapHopper is a bounded model checking framework for Heap-implementations. Overview Setup sudo apt update && sudo apt install build-essential python3-dev virtualenvwrapper git clone https://github.com/angr/heaphopper.git && cd ./heaphopper mkvirtualenv -ppython3 heaphopper pip install ...

NeuPerm

NeuPerm Code for the paper NeuPerm: Disrupting the Extraction of Malware Hidden in Neural Network Parameters by Leveraging Permutation Symmetry. NeuPerm is a zero-cost, zero-retraining sanitization step that disrupts malware hidden in pretrained model weights by reordering permutation-equivalent...

CVE-2026-21636

CVE-2026-21636 - Node.js Permission Model UDS/Network Bypass Vulnerability summary The Node.js permission model --permission is designed to sandbox a process by restricting access to the filesystem, child processes, workers, and network. CVE-2026-21636 reveals that connections made via undici/fet...

open-klara

Open KLara Open KLara is a community-driven fork of KLara by Kaspersky Lab — a distributed system for scanning large malware collections with Yara rules. This project is maintained by the community and is not affiliated with Kaspersky Lab. Features Web Interface — Fire-and-forget Yara scans with...

AI-Threat-Modeler

AI Threat Modeler Point it at a codebase. Get a code-grounded threat model back — ready to review, challenge, and act on. The parts the code can't see — the business context and the architecture behind it — are where you and your architects come in. Upload a ZIP or import a GitHub repo, and AI...

privacy-filter

OpenAI Privacy Filter OpenAI Privacy Filter is a bidirectional token-classification model for personally identifiable information PII detection and masking in text. It is intended for high-throughput data sanitization workflows where teams need a model that they can run on-premises that is fast,...

BrokenHill

!\ Broken Hill logo \https://assets.kitploit.com/production/public/readmes/44608/c14154675c6f33f5c3887887afcdb39351c82f269ee2772cf31847106a20a871.jpg Broken Hill Broken Hill is a productionized, ready-to-use automated attack tool that generates crafted prompts to bypass restrictions in large...

responsible-ai-toolbox

Responsible AI Toolbox Responsible AI is an approach to assessing, developing, and deploying AI systems in a safe, trustworthy, and ethical manner, and take responsible decisions and actions. Responsible AI Toolbox is a suite of tools providing a collection of model and data exploration and...

macOS-Security-and-Privacy-Guide

This guide is a collection of techniques for improving the security and privacy of macOS on Apple silicon Macs. It targets experienced users who want security practices commonly used by organizations, but is also suitable for novice users with an interest in privacy and security. For...

MEA-Bench

MEA-Bench: A Benchmark for Model Extraction Attacks This repository provides a unified benchmark for model extraction attacks, defenses, adaptive attacks, and evaluation. The public interface is organized around a small number of portable commands. Method-specific Python modules and legacy run...

CVE-2024-37054-PoC

Description Proof of concept for CVE-2024-37054, a critical deserialization vulnerability in MLflow versions 0.9.0 to 2.14.1. When a malicious pyfunc model is loaded via mlflow.pyfunc.loadmodel, the pickled payload executes arbitrary code. Files exploit.py — full exploit: register, train, upload,...