AstroGuard: Experimental Results, Execution Logs, and Formal Verification Artifacts

AstroGuard: Experimental Results, Execution Logs, and Formal Verification Artifacts This archive contains the research artifacts generated during the AstroGuard experiments. The project combines machine-learning-based astronomical object analysis with TLA+ formal specifications, TLC model checking, fault-injection testing, and Python reference workflows. The archive includes available experimental datasets, model training outputs, prediction results, evaluation metrics, execution logs, TLA+ specifications, TLC verification reports, reference execution traces, and reproducibility scripts. The experiments include 18K and 60K workflows, classical and quantum-related model experiments, and formal verification scenarios covering release gating, assessment freshness, retry invalidation, and object isolation. The reported reference verification checks completed successfully under their respective configurations. The main TLC configuration disabled deadlock checking, and the Python reference workflows were not verified against a production implementation. Therefore, this archive does not claim production-level safety certification. Source code: https://github.com/Shubha-fm/AstroGuard Version: 1.0.0 Creator: Shubha Chakraborty, Department of Computer Science, The University of Burdwan, India.