figma-to-divi
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 design data from Figma frames, which serves as an ingestion point for untrusted content. Maliciously crafted Figma layers or text nodes could attempt to influence the agent's output during the migration to Divi.
- Ingestion points: Figma node trees and image fills read via the Figma API (SKILL.md).
- Boundary markers: Not explicitly described for separating external design data from internal instructions.
- Capability inventory: Extensive site modification capabilities including 'respira_build_page', 'respira_update_page', 'respira_update_element', and 'respira_sideload_image' (SKILL.md).
- Sanitization: The skill does not mention specific sanitization or filtering of text strings retrieved from Figma before they are processed by the agent.
- [SAFE]: The skill transmits usage telemetry including site version and success status to the developer's official domain at respira.press.
Audit Metadata