design-brief
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input via the "brief" parameter and utilizes the "file_write" capability to create "DESIGN.md" and "brief-preview.html". This pattern constitutes an indirect prompt injection surface where adversarial text in a brief could attempt to manipulate the agent's file-writing behavior.
- Ingestion points: "brief" input (SKILL.md)
- Boundary markers: No explicit delimiters are used to wrap the user-provided brief during processing
- Capability inventory: "file_write" (SKILL.md)
- Sanitization: The skill mitigates risks by requiring the agent to map natural language to a predefined set of symbolic values and concrete design tokens in Sections 1 and 2.1
Audit Metadata