banner-design
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill loads visual style parameters from a local configuration file, creating a surface where instructions could be injected.
- Ingestion points: The instructions require reading brand colors, fonts, and styles from
.agents/design-context.md. - Boundary markers: No specific delimiters or instructions are used to ensure the agent ignores non-data content within the brand file.
- Capability inventory: The skill produces code (HTML/CSS) and image generation prompts, which are influenced by the ingested context.
- Sanitization: The skill does not validate or sanitize the content of the design context file before processing it.
Audit Metadata