XuWY/AgentTrap download history
XuWY/AgentTrap is a text generation dataset on the Hugging Face Hub. In the last 30 days it was downloaded 383 times (44 in the last 7 days), and 1,690 times in total. It ranks #40,700 among datasets by monthly downloads.
AgentTrap AgentTrap is a runtime benchmark for evaluating whether LLM agents can use third-party skills while resisting malicious workflow behavior. This release is generated from the v6 paper-facing dataset. Contents data/tasks.*: the 141-task v6 registry with resolve