product-experiments

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a standard workflow for product experimentation and does not contain any malicious instructions, obfuscation, or unauthorized access patterns.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The instructions explicitly warn against exposing sensitive user data and advise keeping API credentials out of client-side code bundles.
  • [PROMPT_INJECTION]: No attempts to override system prompts or bypass safety guidelines were found.
  • [REMOTE_CODE_EXECUTION]: The skill does not perform remote code execution or download external scripts; it recommends following official documentation for SDK implementation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 06:57 AM
Security Audit — agent-trust-hub — product-experiments