AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security
Published in arXiv preprint, 2026
Core Contributor
AgentDoG 1.5 updates the agent safety taxonomy for Codex and OpenClaw execution scenarios, trains lightweight guardrail variants with a taxonomy-guided data engine, and supports real-time safety moderation in complex interactive agent settings.
Recommended citation: Dongrui Liu, Yu Li, Zhonghao Yang, Peng Wang, Guanxu Chen, Yuejin Xie, Qinghua Mao, Wanying Qu, Yanxu Zhu, Tianyi Zhou, et al. (2026). "AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security." arXiv preprint arXiv:2605.29801. https://arxiv.org/abs/2605.29801