figma-analyze-frame
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill ingests untrusted data (text, layer names, and properties) from external Figma frames via a Model Context Protocol (MCP) server. 1. Ingestion points: Frame data retrieved via the 'Figma Desktop MCP server' from selection-based or link-based inputs. 2. Boundary markers: Absent; there are no instructions to delimit design content from tool instructions. 3. Capability inventory: The skill is limited to generating a markdown analysis report for display. It has no direct file-system write or network exfiltration capabilities. 4. Sanitization: None; text layers from the design are directly interpolated into the implementation analysis.
Audit Metadata