interface-details
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill mentions fetching updated design chapters from the vendor's website at
https://detail.design. This is a legitimate vendor-owned resource for obtaining current guidelines and does not escalate the security verdict.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and modify user-provided UI code, which presents a surface for indirect prompt injection.\n - Ingestion points: Untrusted UI source code and interface definitions are ingested via the 'Build', 'Polish', and 'Review' workflows defined in
SKILL.md.\n - Boundary markers: The skill does not explicitly define delimiters to separate its internal instructions from the user data being analyzed.\n
- Capability inventory: The agent has the capability to modify code and generate reviews, but lacks broader system or network permissions within the skill scope.\n
- Sanitization: There are no explicit instructions for sanitizing or escaping content from the processed UI code.
Audit Metadata