AgentDojo
Emerging9papers using it
2026first seen
AgentDojo is a benchmark dataset used to evaluate the performance and security of tool-using large language model agents in various scenarios.
Papers using AgentDojo (6)
- Adaptive Evaluation of Out-of-Band Defenses Against Prompt Injection in LLM AgentsBeyond Attack-Success Rate: Action-Graded Severity Scale for Tool-Using AI AgentsSecureClaw: Clawing Back Control of LLM AgentsIterInject: Indirect Prompt Injection Against LLM Agents via Feedback-Guided Iterative OptimizationAgentrim: Tool Risk Mitigation For Agentic AIOptimizing Agent Planning for Security and Autonomy