action-view

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is strictly educational, providing documentation and code patterns for Rails developers.
  • [DATA_EXPOSURE]: No hardcoded credentials, API keys, or access to sensitive local file paths (e.g., .ssh, .env) were detected.
  • [REMOTE_CODE_EXECUTION]: There are no patterns involving the download and execution of remote scripts (e.g., curl | bash) or the installation of untrusted packages.
  • [PROMPT_INJECTION]: The content does not contain instructions designed to bypass agent safety filters or override system instructions.
  • [INDIRECT_PROMPT_INJECTION]: While the skill provides templates for handling user-generated content in web views, it correctly documents and demonstrates the use of sanitization helpers like sanitize and strip_tags to prevent XSS. It does not ingest untrusted data into the agent's own execution context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 07:18 AM