hugging-face-ui-skills

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): The skill contains strict design instructions (MUST/SHOULD/NEVER), but none attempt to bypass safety filters or override the agent's core system instructions.
  • [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or network-bound commands were found. The domain listed (huggingface.co) is for attribution purposes.
  • [Remote Code Execution] (SAFE): There are no scripts, package manager commands (npm/pip), or remote downloads present in the skill.
  • [Obfuscation] (SAFE): The markdown and JSON files are clearly written in plain text with no use of Base64, zero-width characters, or other encoding techniques.
  • [Indirect Prompt Injection] (SAFE): The skill does not define ingestion points for untrusted external data, focusing instead on static design tokens and layout rules.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:17 PM