customer-persona
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
infshCLI tool via Bash to execute remote AI applications for market research (Tavily, Exa) and image generation (Fal.ai). - [EXTERNAL_DOWNLOADS]: The documentation suggests the installation of additional tools and skills from the
inference-shrepository using thenpxpackage runner. - [PROMPT_INJECTION]: The skill processes data from external search providers, creating a surface for indirect prompt injection if those sources contain malicious instructions.
- Ingestion points: Market research data and competitive intelligence fetched via
tavily/search-assistantandexa/searchapps. - Boundary markers: No delimiters or boundary markers are defined to separate untrusted search results from the persona generation instructions.
- Capability inventory: Subprocess execution is restricted to the
infshcommand for platform-specific tasks. - Sanitization: There is no evidence of sanitization or filtering of external content before it is interpolated into the persona creation workflow.
Audit Metadata