huggingface-accelerate

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references downloads from official and well-known sources. Specifically, it provides instructions to clone the NVIDIA Megatron-LM and Apex repositories from GitHub for advanced training features.
  • [COMMAND_EXECUTION]: Instructions include standard command-line operations for environment setup (pip install) and training execution (accelerate launch), which are consistent with the primary purpose of distributed training orchestration.
  • [REMOTE_CODE_EXECUTION]: While the skill involves downloading and executing external code (e.g., Megatron-LM), these actions are directed toward well-known, trusted industry repositories and are necessary for the documented functionality.
  • [SAFE]: All code snippets use standard PyTorch and HuggingFace APIs. No patterns of obfuscation, credential harvesting, or unauthorized data access were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 09:58 AM
Security Audit — agent-trust-hub — huggingface-accelerate