tao-train-mask2former

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill fetches pretrained model weights from the SwinTransformer GitHub repository (SKILL.md) and uses official NVIDIA container images (references/tao-deploy-mask2former.md).
  • [INDIRECT_PROMPT_INJECTION]: The skill provides an interface to process user-supplied configuration files and dataset paths for model training and deployment. Ingestion points: user-provided YAML/JSON configuration files and local dataset directory paths (SKILL.md, references/skill_info.yaml). Boundary markers: Absent; user-provided paths are interpolated directly into shell command templates. Capability inventory: Authorized to use the Bash tool to execute CLI commands (SKILL.md). Sanitization: Relies on the underlying TAO Toolkit CLI for configuration validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 02:40 PM
Security Audit — agent-trust-hub — tao-train-mask2former