paperAugust 2026Unreviewed

DreamGuard: Efficient Runtime Guardrail for LLM Agents via Risk-Aware World Model

Wenhao Lin, Chen Yu, Xingwei Lin, Sicong Cao, Xiang Chen, Lei Xue, Le Yu, Letian Sha, Chunming Wu

Abstract

As large language model (LLM) agents increasingly invoke external tools and interact with real-world systems, unsafe actions may cause irreversible consequences on external states, user data, and downstream services. Recent runtime guardrails mitigate such risks by checking proposed actions before execution, but many remain reactive: they primarily assess the apparent safety of the current action, lacking an explicit model of how risk evolves across the trajectory. This limitation creates a crit

Cite

@misc{lin2026dreamguard,
  title = {{DreamGuard: Efficient Runtime Guardrail for LLM Agents via Risk-Aware World Model}},
  author = {Wenhao Lin and Chen Yu and Xingwei Lin and Sicong Cao and Xiang Chen and Lei Xue and Le Yu and Letian Sha and Chunming Wu},
  year = {2026},
  month = aug,
  eprint = {2608.05695},
  archivePrefix = {arXiv},
  url = {https://www.semanticscholar.org/paper/0e72e0b12ecf4eb87226b876a02999df62ca4a33}
}