2026-10-14T00:47:16Z https://www.scidb.cn/oai

10.57760/sciencedb.0106z 2026-08-25T09:36:58Z

2026-08-25 SkyRescue-Bench: A Reproducible Dataset for Runtime Assurance of Safety-Critical Agentic Dynamic Workflows doi:10.57760/sciencedb.0106z en kyRescue-Bench is a reproducible experimental dataset supporting the paper SkyRescue: Multi-Agent Dynamic Workflow Compilation, Orchestration, and Runtime Assurance for Safety-Critical Agent Systems. The dataset focuses on low-altitude emergency traffic command scenarios and covers runtime assurance mechanisms for safety-critical agent systems, including untrusted language candidates, symbolic grounding, typed intent compilation, external side-effect commit protocols, and commitment-preserving local repair. It provides anonymized human instructions, the HeldOut100 gold set, frozen raw model responses, experimental result files, 2,000 synthetic stress-test tasks generated with fixed random seeds, dataset documentation, and SHA-256 checksum manifests.This dataset is intended to support the reproduction of the experimental analyses and ablation studies reported in the paper. It enables offline verification of metrics such as schema parsing, semantic field accuracy, entity grounding, safety decision making, runtime event repair, commitment preservation, duplicate-call prevention, and change ratio. The dataset does not contain real flight operation logs, expert identity information, API keys, or personally identifiable information. The synthetic stress-test data are provided solely for research evaluation and should not be interpreted as evidence of real airspace operation or actual flight-control deployment. Safety-critical agents; Dynamic workflow; Runtime assurance; Symbolic grounding; Low-altitude emergency traffic; Multi-agent systems; Reproducible evaluation; Synthetic benchmark dataset Liu Yushu EMBARGO https://creativecommons.org/licenses/by-nc-sa/4.0/ dataset Science Data Bank