Building a Foundational Guardrail for General Agentic Systems via Synthetic Data
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912642370633728 |
|---|---|
| author | Huang, Yue Hua, Hang Zhou, Yujun Jing, Pengcheng Nagireddy, Manish Padhi, Inkit Dolcetti, Greta Xu, Zhangchen Chaudhury, Subhajit Rawat, Ambrish Nedoshivina, Liubov Chen, Pin-Yu Sattigeri, Prasanna Zhang, Xiangliang |
| author_facet | Huang, Yue Hua, Hang Zhou, Yujun Jing, Pengcheng Nagireddy, Manish Padhi, Inkit Dolcetti, Greta Xu, Zhangchen Chaudhury, Subhajit Rawat, Ambrish Nedoshivina, Liubov Chen, Pin-Yu Sattigeri, Prasanna Zhang, Xiangliang |
| contents | While LLM agents can plan multi-step tasks, intervening at the planning stage-before any action is executed-is often the safest way to prevent harm, since certain risks can lead to severe consequences once carried out. However, existing guardrails mostly operate post-execution, which is difficult to scale and leaves little room for controllable supervision at the plan level. To address this challenge, we highlight three critical gaps in current research: data gap, model gap, and evaluation gap. To close the data gap, we introduce AuraGen, a controllable engine that (i) synthesizes benign trajectories, (ii) injects category-labeled risks with calibrated difficulty, and (iii) filters outputs via an automated reward model, producing large and reliable corpora for pre-execution safety. To close the guardian model gap, we propose a foundational guardrail Safiron, combining a cross-planner adapter with a compact guardian model. The adapter unifies different input formats, while Safiron flags risky cases, assigns risk types, and generates rationales; trained in two stages with a broadly explored data recipe, Safiron achieves robust transfer across settings. To close the evaluation gap, we release Pre-Exec Bench, a realistic benchmark covering diverse tools and branching trajectories, which measures detection, fine-grained categorization, explanation, and cross-planner generalization in human-verified scenarios. Extensive experiments demonstrate consistent gains of the proposed guardrail over strong baselines on Pre-Exec Bench, and ablations further distill actionable practices, providing a practical template for safer agentic systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_09781 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Building a Foundational Guardrail for General Agentic Systems via Synthetic Data Huang, Yue Hua, Hang Zhou, Yujun Jing, Pengcheng Nagireddy, Manish Padhi, Inkit Dolcetti, Greta Xu, Zhangchen Chaudhury, Subhajit Rawat, Ambrish Nedoshivina, Liubov Chen, Pin-Yu Sattigeri, Prasanna Zhang, Xiangliang Machine Learning Artificial Intelligence Computation and Language While LLM agents can plan multi-step tasks, intervening at the planning stage-before any action is executed-is often the safest way to prevent harm, since certain risks can lead to severe consequences once carried out. However, existing guardrails mostly operate post-execution, which is difficult to scale and leaves little room for controllable supervision at the plan level. To address this challenge, we highlight three critical gaps in current research: data gap, model gap, and evaluation gap. To close the data gap, we introduce AuraGen, a controllable engine that (i) synthesizes benign trajectories, (ii) injects category-labeled risks with calibrated difficulty, and (iii) filters outputs via an automated reward model, producing large and reliable corpora for pre-execution safety. To close the guardian model gap, we propose a foundational guardrail Safiron, combining a cross-planner adapter with a compact guardian model. The adapter unifies different input formats, while Safiron flags risky cases, assigns risk types, and generates rationales; trained in two stages with a broadly explored data recipe, Safiron achieves robust transfer across settings. To close the evaluation gap, we release Pre-Exec Bench, a realistic benchmark covering diverse tools and branching trajectories, which measures detection, fine-grained categorization, explanation, and cross-planner generalization in human-verified scenarios. Extensive experiments demonstrate consistent gains of the proposed guardrail over strong baselines on Pre-Exec Bench, and ablations further distill actionable practices, providing a practical template for safer agentic systems. |
| title | Building a Foundational Guardrail for General Agentic Systems via Synthetic Data |
| topic | Machine Learning Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2510.09781 |