timesfm-forecasting

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates downloading pretrained model weights from Hugging Face (e.g., google/timesfm-2.5-200m-pytorch) and installing dependencies via official Python registries. These sources are recognized as trusted technology services.
  • [COMMAND_EXECUTION]: The scripts/check_system.py utility utilizes subprocess.run to query system statistics like RAM availability using the sysctl and vm_stat commands. These operations are benign diagnostic checks intended to prevent system instability during model loading. The utility also uses dynamic module importing via __import__ to verify that essential libraries like torch and timesfm are correctly configured in the environment.
  • [PROMPT_INJECTION]: The skill instructions use 'CRITICAL' markers to emphasize the importance of hardware preflight checks. These are legitimate operational warnings regarding resource usage and do not constitute attempts to override core agent behavior or security constraints.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 08:22 AM
Security Audit — agent-trust-hub — timesfm-forecasting