List your company

Tool-invocation hijacking to RCE on Agent Security Scenarios New LLM Lineup

100Gemini-3.1-pro RCE Success RateRCE-2
6.430.75579.3Sep 6, 2025

Evaluation results

MethodLinks
RCE-2Agent=Cline, Agent Version=New2025.09100100——20—
RCE-2Agent=Trae, Agent Version=New2025.09100—————
RCE-2Agent=Cursor, Agent Version=New2025.0930—————
AgentDojoAgent=Cline, Agent Version=New2025.0910—————
RCE-2Agent=WindSurf, Agent Version=New2025.0910——100——