用 Jev 与 CodeGraph 组合做安全审查
开发者分享了 typesafe-security-review 项目:把 Jev 的概率判断与代码图、OWASP/CWE 信息结合,用于低成本安全扫描。
对800个提示的测试显示,不同提示注入分类器在不同的攻击类型上失败。Jev的漏报集中在技术混淆,Clef在社交工程,Wolf在语言技巧和上下文操纵。Horizon Guard仅漏掉5个(但...)。
Prompt-injection classifiers don't just differ in accuracy. They break on different attacks. Across 800 prompts, Jev misses cluster in technical obfuscation, Clef in social engineering, and Wolf's in linguistic tricks and context manipulation. Horizon Guard misses only 5 (but it