figma-to-swiftui
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external design data from Figma, representing an attack surface for indirect prompt injection.
- Ingestion points: Design context, metadata, and asset URLs retrieved from the Figma MCP server (identified in SKILL.md and references/fetch-strategy.md).
- Boundary markers: Instructions suggest treating web code as 'design evidence' (SKILL.md), but no explicit prompt boundary markers are present to isolate ingested content.
- Capability inventory: Capabilities include writing code files and executing shell commands (references/asset-handling.md).
- Sanitization: There is no mention of sanitization for design data text.
- [COMMAND_EXECUTION]: The skill utilizes system shell commands for image asset management.
- Evidence: Commands such as
sipsfor image resizing andcpfor asset organization are referenced in references/asset-handling.md. These are standard utilities for the task.
Audit Metadata