member-insights
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [NO_CODE] (SAFE): The provided content consists entirely of markdown documentation. It does not include any executable scripts (.py, .js, .sh), binary files, or tool definitions that an agent could execute.
- [DATA_EXPOSURE] (SAFE): While the documentation mentions sensitive data sources such as transaction systems and CRM databases, it does not provide methods to access these systems, nor does it contain hardcoded credentials or API keys.
- [PROMPT_INJECTION] (SAFE): The text contains standard instructional language and framework templates. No patterns for system prompt overrides, jailbreaks, or instruction-ignoring behaviors were detected.
- [INDIRECT_PROMPT_INJECTION] (SAFE): The skill identifies external data sources (surveys, support tickets) as part of its framework but provides no functional implementation or code to process this data, meaning there is no active attack surface for indirect injection.
Audit Metadata