figma-design-sync

Pass

Audited by Gen Agent Trust Hub on Mar 27, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the agent-browser CLI to navigate to web URLs, take snapshots, and capture screenshots (agent-browser open, agent-browser snapshot, agent-browser screenshot). This is a core part of its visual comparison functionality.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by processing content from external sources.
  • Ingestion points: Figma design specifications (via Figma MCP) and web page content/metadata (via agent-browser).
  • Boundary markers: None identified. The agent is instructed to extract and analyze design properties directly from these external environments.
  • Capability inventory: The skill has permissions to modify local source code (CSS, Tailwind, HTML/ERB files) and execute browser-based commands.
  • Sanitization: No explicit validation or sanitization logic is present to filter instructions that might be embedded in Figma text nodes or web page content.
  • [SAFE]: All identified behaviors—including external network access to Figma, browser automation, and local file modification—are consistent with the primary purpose of a design-sync agent. The skill uses recognized tools and follows standard developer workflows.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 27, 2026, 05:25 PM