腾讯云智能渗透黑客松 Official repository of Tencent Cloud Intelligent Penetration Hackathon. Showcasing top open-source projects of LLM-based autonomous penetration agents, including multi-agent collaboration, automated penetration, AI-driven offensive security, and intelligent attack-defense solutions.
Red Team AI Benchmark: Evaluating LLMs for authorized offensive-security tasks
An authorization-first, policy-governed red-team operations platform that turns scoped security testing into safe, auditable, and evidence-ready workflows
Your AI agent has no idea it's been hijacked. AgentGuard does
Runtime security fabric for AI agents. Policy-as-code governance with trust segmentation, cognitive velocity detection, behavior chain analysis, and multi-agent identity tracking.
Autonomous Purple Teaming framework combining agentic penetration testing with dynamic, real-time SOC rule synthesis.
To associate your repository with the offensive-ai topic, visit your repo's landing page and select "manage topics."