#1LLM security, prompt injection, model extraction, adversarial AI, and AI red teaming tools.
Kitploit recommended

ENDGAME C2 FRAMEWORK — AI-powered command and control for professional red team operations
CybORG CAGE 2 environment updated for the paper "Alignment of Autonomous Cyber Defense Agents via Adaptive Policy Shaping"

Privacy testing library for deep learning systems, enabling assessment of susceptibility to membership inference, model extraction, and model…

A collection of real world AI/ML exploits for responsibly disclosed vulnerabilities

Agentic AI memory with Ebbinghaus forgetting curve decay. +16pp better recall than Mem0 on LoCoMo.

Malware Mutation Using Reinforcement Learning and Generative Adversarial Networks

LLM-agent-powered concolic execution engine that instruments source code, summarizes path constraints in natural language, and generates test cases…

NOVA - Claude Code Protection System against prompt injection attacks

OWASP-aligned intentionally vulnerable platform for learning and testing AI, LLM, RAG, MCP and agentic AI security

AIO Cloud Managment Server

Pre-install security for AI agents, npm packages, and MCP servers. Zero-dep local static analysis; normal scans never execute package code.

Research implementation of the Black-Hole Attack that poisons vector database indexes via centrality-driven hubness, with FAISS pipelines and…

Security scanner auditing Claude Code environments for CVE-2026-21852 pre-trust execution, hook hijacking, and eBPF lockdown.

AI Smart Contract Security Analysis and PoC Generation Framework

Pre-execution intent verification for AI agents. Audits what your AI is about to do, not what it says. Zero dependencies, deterministic, hash-sealed.

Research code and dataset (WSD) for evaluating audio watermarking impact on anti-spoofing, using wav2vec2 XLS-R and knowledge-preserving learning.

Automated threat hunting pipeline that ingests Azure logs, uses LLM reasoning to detect suspicious activity, assess risk, and generate remediation…

PyTorch reference implementation of a fast adversarial training method that prevents catastrophic overfitting via coupled objectives and an EMA-based…