展示 AI Agent 做漏洞利用的能力边界,对自建 Agent 做安全任务有直接参照。
ai_agentlpe内核
RT @Jhaddix: Big thanks to our sponsor this week, @Xbow!
Really cool breakdown from the @Xbow team on finding a Linux kernel vulnerability, turning it into a working LPE exploit, learning about where autonomy still breaks down and how the team solved it.
https://t.co/SduWz5d41o