image-to-code
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided design requests and content, creating an attack surface where malicious instructions could be embedded in the design requirements to influence the generated code or agent behavior.
- Ingestion points: The skill ingests untrusted data from user prompts that describe website sections, branding, and content (e.g., 'Example Interpretations' in SKILL.md).
- Boundary markers: There are no explicit delimiters or boundary markers defined in the skill to separate user-provided content from the agent's internal instructions.
- Capability inventory: The skill instructs the agent to perform front-end implementation (HTML/CSS/JS generation), which is the primary capability that could be targeted by an injection attack.
- Sanitization: The instructions include qualitative guidelines to avoid 'AI slop' and 'fake brand slop,' but lack programmatic sanitization or escaping of the user-provided content before it is interpolated into the design or implementation process.
Audit Metadata