Adversarial AI bug hunter with auto-fix skill for Claude Code, Cursor, Codex CLI, GitHub Copilot CLI, Kiro CLI, Opencode, Pi Coding Agent, and more. Multi-agent pipeline finds security vulnerabilities, logic errors, and runtime bugs — then fixes them autonomously on a safe branch.

Semantic Stealth Attacks & Symbolic Prompt Red Teaming on GPT and other LLMs.

VEX Protocol — The trust layer for AI agents. Adversarial verification, temporal memory, Merkle audit trails, and tamper-proof execution. Built in Rust.

AATMF | An Open Source - Adversarial AI Threat Modeling Framework

Basilisk — Open-source AI red teaming framework with genetic prompt evolution. Automated LLM security testing for GPT-4, Claude, Gemini. OWASP LLM Top 10 coverage. 32 attack modules.

Test and evaluate Large Language Models against prompt injections, jailbreaks, and adversarial attacks with a web-based interactive lab.

LLM Attack Testing Toolkit is a structured methodology and mindset framework for testing Large Language Model (LLM) applications against logic abuse, prompt injection, jailbreaks, and workflow manipulation.

Audit legacy codebases with adversarial AI agent teams — 7 iterations, 168 findings, 81.8% reliability score

Proof of concept tool to bypass document replay technology (such as GPTZero).

Implementation of Vocabulary-Based Adversarial Fuzzing (VB-AF) to systematically probe vulnerabilities in Large Language Models (LLMs).

🛡️ Enterprise-grade AI security framework protecting LLMs from prompt injection attacks using ML-powered detection

LLM Sentinel Red Teaming Platform is an enterprise-grade framework for automated security testing of Large Language Models, detecting vulnerabilities such as jailbreaks, prompt injection, and system prompt leakage across multiple providers, with structured attack orchestration, risk scoring, and security reporting to harden models before production

A research framework for simulating, detecting, and defending against backdoor loop attacks in LLM-based multi-agent systems.

Breaking Chain-of-Thought: A Comprehensive Taxonomy of Reasoning Vulnerabilities in Production AI Systems

Slides and materials from cybersecurity talks at Chubut Hack (2021-2022)

1st Place Winner (General Judge) - Datadog Self-Improving Agents Hack. Two identical AI agents play Split or Steal. No pre-programmed betrayal. They discover deception on their own. Built with @evancorrea.

A complete self-hosted AI research platform running on Docker with GPU acceleration. Combines LLM inference, vector search, web search, code execution. and fully searchable logging with Splunk - all running locally.

Pit AI models against each other. Score them sealed. Crown a winner. All built using the GitHub Copilot CLI. ⚡

🔍 Emulate advanced phishing tactics ethically with this open-source framework for red team operations focused on social engineering sophistication.

👻 Adversarial AI Pentester - CHAOS vs ORDER dual-agent exploitation with collective memory