graphic-design
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXPOSURE]: The skill reads project-specific configuration from
.agents/design-context.md. This is a legitimate use of local context files for maintaining brand consistency and does not target sensitive system credentials or private user data. - [EXTERNAL_DOWNLOADS]: The skill generates HTML templates that include links to Google Fonts (
fonts.googleapis.com). This is an expected reference to a well-known, trusted service for web typography. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data (user design briefs and product images) which are then used to construct AI prompts and HTML output.
- Ingestion points: User-provided text briefs and product image attributes (SKILL.md).
- Boundary markers: Absent; there are no specific delimiters defined to separate user input from system instructions.
- Capability inventory: The skill is limited to generating static code (HTML/CSS) and text prompts; it lacks capabilities for command execution, network exfiltration, or file modification.
- Sanitization: Absent; the skill relies on the underlying LLM's safety filters when generating prompts and code.
Audit Metadata