brand-storytelling
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface (Category 8) because it processes untrusted user data as part of its brand storytelling workflow.\n
- Ingestion points: User-supplied information about their product and origin story enters the context through prompts and intake responses defined in
references/INTAKE.md.\n - Boundary markers: The skill does not employ specific delimiters or 'ignore embedded instructions' warnings for the user-supplied brand material.\n
- Capability inventory: Capabilities are limited to generating narrative text and channel-specific scripts in markdown format; there are no tools for file system modification, shell access, or external communication.\n
- Sanitization: No input sanitization or validation is performed on the raw story material provided by the user.
Audit Metadata