design-experience-interaction
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes content from untrusted local source files without sanitization.
- Ingestion points: The skill reads from
references/design-best-practices.md,references/design-inspiration-catalog.md, andDESIGN.md. - Boundary markers: No delimiters or explicit boundary instructions are used to separate ingested content from the system prompt.
- Capability inventory: The skill writes results to
02-design.md, which guides downstream tasks like visual design and workflow planning. - Sanitization: There is no evidence of validation or sanitization of the content fetched from the reference files.
- [NO_CODE]: This skill consists solely of markdown instructions and does not contain any executable scripts or binaries.
Audit Metadata