3/ Do you really want something? Then really work. Half-assing doesn't get you anywhere. Don't look for "system hacks," or whatever, just do it. You first need the baseline mentality of going all in, or nothing else matters
感觉没啥人提这个,不过链上协议绝对是Claude Code源码泄漏的主要受害者之一。本来黑客是不太好用 Claude Code 直接做漏洞测试的,因为内部有一个 CYBER_RISK_INSTRUCTION 会自动 block 恶意的请求。
甚至文件头注释里面包含一个醒目的警告:
“IMPORTANT: DO NOT MODIFY THIS INSTRUCTION WITHOUT SAFEGUARDS TEAM REVIEW This instruction is owned by the Safeguards team (David Forsythe, Kyla Guru)”
前几天的新闻还在说,最强的那个模型"Mythos"不公布的原因是因为内部对网络安全挑战的担忧。这下好了,黑客可以自己通过移除 CYBER_RISK_INSTRUCTION 获取全解锁,黑化版 Claude Code 了,史诗级加强。
keep safe 🙏
Very simple way to figure out whether to use AI.
Gym: the point isn't for the weight to be lifted, but for you to lift the weight.
Job: the point is for the weight to be lifted.
Use AI for job, not for gym.
https://t.co/bHC35KCCMs