18-referral-program-global

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns detected. The skill consists entirely of markdown documentation, templates, and legal compliance guidelines for marketing operations.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: No network calls, hardcoded credentials, or sensitive file access patterns were found. The skill suggests reading a local marketing context file which is a standard operational practice.
  • [REMOTE_CODE_EXECUTION]: The skill does not perform any external downloads, package installations, or remote script executions.
  • [PROMPT_INJECTION]: The instructions are clear and instructional without any attempts to override safety filters or disregard system rules.
  • [DYNAMIC_EXECUTION]: No code generation or runtime execution of dynamic content is present.
  • [OBFUSCATION]: No encoded content, zero-width characters, or homoglyph attacks were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 11:39 PM
Security Audit — agent-trust-hub — 18-referral-program-global