agency-retail-customer-returns
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of instructional text and persona guidelines. No executable scripts, network operations, or external dependencies are present.
- [PROMPT_INJECTION]: Instructions include guidelines for the agent's behavior (e.g., 'Critical Rules You Must Follow'), but these are standard for defining a professional persona and do not attempt to bypass safety filters or override the underlying system prompt.
- [DATA_EXFILTRATION]: The skill references customer data (names, transactions, credit cards) exclusively within the context of a simulated retail role-play. There are no mechanisms for actual data exfiltration or hardcoded credentials.
- [INDIRECT_PROMPT_INJECTION]: While the skill is designed to process user-provided information (return reasons, customer stories), it lacks the capabilities (such as file writing or network access) that would make it vulnerable to exploitation via indirect injection.
Audit Metadata