评分 7 · 来源：cs.AI updates on arXiv.org · 发布于 2026-04-14
评分依据：Addresses critical attack vector: indirect prompt injection via tool-returned content. Three primary attack surfaces covered. Essential security infra for tool-using agents.
ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection
发布
采集
学术前沿 7.0 分
— Addresses critical attack vector: indirect prompt injection via tool-returned content. Three primary attack surfaces covered. Essential security infra for tool-using agents.