implementing-llms-litgpt

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation for the 'litgpt' library, which is a reputable open-source project by Lightning AI. All mentioned dependencies (torch, transformers) are standard industry packages.
  • [EXTERNAL_DOWNLOADS]: The skill references downloading model weights from Hugging Face Hub (e.g., meta-llama, microsoft, google). These are well-known services and trusted model providers. The download operations are performered via the legitimate 'litgpt download' CLI tool.
  • [DATA_EXPOSURE]: While the skill mentions using 'HF_TOKEN' for gated models, it correctly instructs the user to export it as an environment variable rather than hardcoding it, which aligns with security best practices for credential management.
  • [COMMAND_EXECUTION]: The skill uses standard CLI commands for machine learning workflows (pip, python, litgpt, sbatch). These are consistent with the skill's stated purpose of training and deploying models.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:07 PM
Security Audit — agent-trust-hub — implementing-llms-litgpt