ChainReactor is a research project that leverages AI planning to discover exploitation chains for privilege escalation on Unix systems. The project models the problem as a sequence of actions to achieve privilege escalation from initial access to a target system.

Abuse suid mis-configured binaries like a pro with -4buzer.

🪲 Bikōchū (微香虫) - DIY Self-Hosted Shodan-Like. Remember, track, uncover: The shinobi way of Bikōchū's scanning.

This llm guardrail is an open-source, dual-layer AI input/output guardrail application designed to secure downstream Large Language Models (LLMs) against malicious attacks and data leaks. Built with Streamlit and Anthropic, the application actively intercepts both user inputs and model responses to ensure safe and compliant interactions.