figma-build

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, including UI descriptions and various code formats (React, Vue, Tailwind), which can be used as a vector for indirect prompt injection.
  • Ingestion points: Untrusted data enters the agent context via user-provided UI descriptions and source code snippets as described in SKILL.md and references/assemble-screens.md.
  • Boundary markers: The instructions do not specify the use of clear delimiters or instructions to ignore embedded commands within the source data.
  • Capability inventory: The skill utilizes a wide range of Figma-specific capabilities, such as creating frames, importing SVG/images, and managing design system variables across all referenced files, but does not use general-purpose high-risk tools like shell command execution or external network requests.
  • Sanitization: There are no instructions for sanitizing, validating, or filtering the content of the source code or descriptions before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:35 AM
Security Audit — agent-trust-hub — figma-build