The Agent Skills Directory

[SAFE]: The skill provides legitimate instructional content for the nvalchemi library, covering domain decomposition strategies and multi-GPU simulation scaling.
[SAFE]: Code snippets use standard PyTorch and nvalchemi APIs for distributed computing (e.g., DistributedManager, DomainParallel). Use of torchrun and SLURM is expected for high-performance computing environments.
[SAFE]: Model loading via MACEWrapper.from_checkpoint refers to standard model identifiers. No suspicious external URLs or untrusted remote code execution patterns were found.
[SAFE]: No evidence of prompt injection, obfuscation, or unauthorized access to sensitive files (e.g., SSH keys, credentials) was detected.

nvalchemi-distributed