simpo-training

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • EXTERNAL_DOWNLOADS (SAFE): The skill references and clones code from the huggingface GitHub organization, which is a verified and trusted source.
  • COMMAND_EXECUTION (SAFE): Training workflows use standard tools like conda, pip, and accelerate to manage environments and execute scripts from trusted local paths.
  • REMOTE_CODE_EXECUTION (SAFE): Code execution is limited to scripts from the trusted alignment-handbook repository. No high-risk patterns such as piping remote URLs directly into a shell were observed.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:22 PM