ai-infra

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill implements standard operational procedures for Azure HPC infrastructure management, focusing on GPU performance and cluster health.
  • [COMMAND_EXECUTION]: Utilizes verified system tools like nvidia-smi, ibstat, and scontrol for diagnostic and management tasks.
  • [SAFE]: Accesses Azure-specific metadata via IMDS and Hyper-V KVP interfaces to gather necessary identifiers for support requests.
  • [SAFE]: Interacts with official Azure Management API endpoints to facilitate the Guest Health Reporting workflow for hardware remediation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 07:30 AM
Security Audit — agent-trust-hub — ai-infra