specialist-design-ui-designer
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interpolates untrusted user input directly into a prompt for a sub-agent without using boundary markers or sanitization. This creates a surface for indirect prompt injection where a user's question could potentially influence the behavior of the sub-agent. \n- Ingestion points: User input enters the context through the
{the user's question}placeholder inSKILL.md. \n- Boundary markers: No delimiters or 'ignore embedded instructions' warnings are present around the user input. \n- Capability inventory: The skill utilizes theAgenttool to delegate tasks to an ephemeral specialist sub-agent. \n- Sanitization: No escaping or validation of the input is performed before interpolation.
Audit Metadata