AI-native red-team for penetration testing and vulnerability research

The AI-native red-team project provides a workbench for authorized penetration testing and vulnerability research, using specialist AI agents, sandboxed tools, evidence records, and replayable timelines. It is designed for security researchers and ethical hackers to automate and scale complex attack simulations with greater precision and auditability. Its use of replayable timelines and agent-driven workflows makes it interesting for advancing autonomous, reproducible security testing.