timesfm-forecasting

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads pretrained model weights (safetensors) from Google Research's official repository on the Hugging Face Hub. This is a standard and expected behavior for a foundation model implementation.
  • [COMMAND_EXECUTION]: The scripts/check_system.py utility executes system commands such as sysctl and vm_stat on macOS or reads /proc/meminfo on Linux to verify physical memory and GPU availability. These operations are performed to ensure the local environment has sufficient resources before attempting to load the large model into memory.
  • [DYNAMIC_EXECUTION]: The preflight check uses dynamic importing to verify that the required Python dependencies (timesfm, torch) are installed and functional before proceeding with the main forecasting logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 10:21 AM