Curated security vulnerability benchmarks & vulnerable code samples for evaluating SAST tools and LLMs. Java · Node.js · Go · Next.js · CVE datasets · 2025-2026 research. OWASP Benchmark, NIST Juli…

Autonomous CTF agents for DEF CON 34 / AI Village HalCTF competition

A collection of public writeups on attacks against AI agents: prompt injection, tool abuse, memory poisoning, and data exfiltration.

The curated list of tools, standards, and resources for governing AI agents in production: policy enforcement, audit trails, EU AI Act compliance, MCP and Claude Code security.

The agent that grows with you

Prompt Hardener analyzes prompt-injection-originated risk in LLM-based agents and applications.

A growing collection of solutions, tools, protocols and techniques to help secure where agents run

LLM-engineering-nlp is a production-focused repository for Advanced NLP and Generative AI engineering. It features hands-on implementations of transformers, large language models (LLMs), fine-tunin…

AgentLeak: Open benchmark for privacy leakage in LLM agents — 7 channels, multi-agent, multi-framework.

🎨 Best DeepSeek Harness Design Plugin. The open-source Claude Design alternative. 🖥️ Local-first desktop app. 🖼️ Your coding agent becomes the design engine: prototypes, landing pages, dashboards, …

Demos for Practical AI Security by No Starch Press

Security middleware for x402 agentic payments — PII redaction, spending policy, and replay detection before blockchain commit

Collection of evals for Inspect AI

A curated collection of AI agent research papers released in 2026, covering agent engineering, memory, evaluation, workflows, and autonomous systems.

DynAuditClaw — A security audit skill that dynamically discovers your OpenClaw agent's real configuration, designs targeted attack scenarios adapted to your specific setup, and executes them in iso…

A Dynamic Environment to Evaluate Attacks and Defenses for LLM Agents.

The official implementation of the paper "AgentDyn: Are Your Agent Security Defenses Deployable in Real-World Dynamic Environments?"

Practice created for AI Security course

universal text transformation, translation, mutation, encoding/decoding, and promptcrafting

This repository contains the Hugging Face Agents Course.

Red Teaming python-framework for testing chatbots and GenAI systems.

ArgFuscator.net is an open-source, stand-alone web application that helps generate obfuscated command lines for common system-native executables.

Papers about red teaming LLMs and Multimodal models.