design-md-maker
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes external design references and project UI evidence to generate DESIGN.md documentation, which creates an attack surface where malicious instructions could be embedded in the processed data. 1. Ingestion points: The skill reads local project files including UI source, theme files, and user-provided external URLs referenced in SKILL.md and rules/project-design-discovery.md. 2. Boundary markers: The instructions emphasize treating retrieved content as evidence rather than authority and include validation checks for token references in rules/validation.md. 3. Capability inventory: Tool access includes read, search, find, edit, write, and URL reads. 4. Sanitization: There is no explicit content sanitization or escaping mechanism defined for the ingested data.
- [DATA_EXFILTRATION]: The skill performs network requests to retrieve design information from non-whitelisted domains such as getdesign.md. These operations are used for documentation discovery and are not associated with sensitive file access.
- [DATA_EXFILTRATION]: The skill fetches design guidelines and metadata from Google's official documentation. This is a reference to a well-known service provider.
Audit Metadata