
ezEmu
See adversary, do adversary: Simple execution of commands for defensive tuning/research (now with more ELF on the shelf)

See adversary, do adversary: Simple execution of commands for defensive tuning/research (now with more ELF on the shelf)

Open-weight LLMs and training/evaluation code for defending against prompt injection attacks, with benchmarks for agentic tool-calling and…

Generates HTML smuggling pages that embed and reconstruct files client-side via JavaScript, with payload encoding, chunking, obfuscation, and…

Benchmark and evaluation harness testing whether LLM agents resist malicious instructions hidden in multimodal skill images, with 108 cases across…

Open framework for RL-based prompt injection red teaming, with a shared trainer, curriculum learning, and benchmarks like AgentDojo, InjecAgent, and…

Benchmark for evaluating AI agent safety against attacks embedded in skill-facing context, with 155 cases across 6 risk domains, measuring task…

A high-severity prompt injection flaw in Claude AI proves that even the smartest language models can be turned into weapons — all with a few lines of…

PoC repository for the blog post CopyEscape: Taking Over Docker Hosts with docker cp

Hands-on AI security lab platform with 50+ scenarios across prompt injection, agentic system exploitation, model manipulation, and MCP trust boundary…

Web-based adversary emulation platform that orchestrates Atomic Red Team tests across Windows endpoints via Go agents, with MITRE ATT&CK mapping, APT…

Security research on context compaction in coding agents: shows how forged role labels in untrusted content survive summarization and get executed,…

CEREBRO-RED v2: Advanced LLM Red Team Research Platform with PAIR Algorithm and LLM-as-a-Judge Evaluation

Research implementation for mitigating adaptive prompt injections via on-policy distillation, with training recipes and evaluators for SEP, PISmith,…

Security benchmark for evaluating OpenClaw agents against adversarial execution contexts including poisoned files, injected skills, misleading tool…

Reverse Shell Detection with Machine Learning

Practical black-box adversarial packet generation against encrypted traffic classification with minimal overhead and full packet recoverability.

Red-teaming benchmark and black-box attack framework for evaluating visual misinformation risks in commercial image generators, with 10k false-claim…

Benchmarking framework for evaluating computer-use AI agents against multi-step indirect prompt injection, with automatic adversarial goal…