paper-ds-land
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from design revisions and artwork, which presents a surface for indirect instructions.
- Ingestion points: Reads design subjects, revisions, artwork, and source code files.
- Boundary markers: No explicit delimiters or ignore-instructions blocks are used for ingested design content.
- Capability inventory: Modifies component pages and metadata mappings; executes extraction tools like get_jsx.
- Sanitization: No validation or sanitization of external design content is documented.
- [SAFE]: No malicious patterns, suspicious network activities, or attempts at unauthorized code execution were identified. The workflow is focused on legitimate design system maintenance and requires user approval for significant changes.
Audit Metadata