skills/nvidia/nemo-fabric/maintain-ci/Gen Agent Trust Hub

maintain-ci

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill is designed to enhance security by enforcing the use of minimum permissions and pinning third-party GitHub Actions to specific commit SHAs.
  • [COMMAND_EXECUTION]: Employs standard local tools including just for formatting checks and ripgrep (rg) to inspect workflow configurations. These operations are limited to development validation tasks.
  • [SAFE]: External references to GitHub Actions such as astral-sh/setup-uv and Swatinem/rust-cache are standard, well-known ecosystem utilities used for their intended purposes.
  • [SAFE]: The skill correctly advises developers to manage secrets by isolating them to specific jobs rather than hardcoding them or granting broad access.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 08:01 AM
Security Audit — agent-trust-hub — maintain-ci