Detecting Prompt Injection Attacks on Purpose-Specific LLM Agents

2 points | by zhinit 5 hours ago