customer-persona

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes data retrieved from external search providers to build personas.
  • Ingestion points: Market research data is ingested through the tavily/search-assistant, exa/search, and exa/answer apps as defined in SKILL.md.
  • Boundary markers: The instructions do not define specific delimiters to separate untrusted search results from the persona generation prompts.
  • Capability inventory: The skill is restricted to the belt CLI tool, which is used to run specific research and image generation applications.
  • Sanitization: There is no explicit sanitization of external search data before it is incorporated into the persona templates.
  • [EXTERNAL_DOWNLOADS]: The skill references external resources for installation and extended functionality.
  • Provides links to installation guidelines hosted on GitHub's raw content service.
  • Recommends the installation of the belt CLI and related skills via npx.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 02:18 PM
Security Audit — agent-trust-hub — customer-persona