run-on-slurm

Pass

Audited by Gen Agent Trust Hub on May 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security risks were detected. The skill provides technical instructions for configuring and launching distributed training jobs on SLURM clusters using standard tools like sbatch and srun.
  • [COMMAND_EXECUTION]: The skill provides a bash template for job submission. This involves standard cluster operations for multi-node training and does not include suspicious or obfuscated commands.
  • [EXTERNAL_DOWNLOADS]: The skill mentions environment setup using 'uv', which involves downloading packages from standard registries. This is an expected and safe practice for preparing the training environment.
Audit Metadata
Risk Level
SAFE
Analyzed
May 3, 2026, 07:06 AM