paper-ds-options
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection via external design data from Figma and Mobbin. While a risk factor, the skill includes sanitization logic for the data it processes.
- Ingestion points: The skill reads design tokens, component variants, and layer names from Figma (references/figma.md) and search results/images from Mobbin (references/references.md).
- Boundary markers: There are no explicit delimiters used to separate external design metadata from agent instructions.
- Capability inventory: The skill uses paper-ds-tokens to write design system tokens and paper-ds-components to draw UI elements on the canvas.
- Sanitization: The skill resolves Figma aliases to literal CSS values and filters instance node IDs to their last segment before processing, reducing the risk of malicious structural injection.
- [EXTERNAL_DOWNLOADS]: The skill integrates with external design platforms to fetch resources.
- Evidence: It utilizes Figma and Mobbin MCP servers to import design assets and references as requested by the user. These integrations use well-known design services and are core to the skill's legitimate design workflow.
Audit Metadata