figma-agent

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). Runtime path reads user-supplied Figma file/node data (parsed from a pasted figma.com link) via figma-mcp-server tools like get_file_nodes, render_images, and token/variable extraction, so outsider-authored free text inside the Figma file (e.g., text layers and comments) is ingested.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs installing and running remote code via the command that fetches from https://github.com/planetabhi/design-diff (npx skills add ...), which downloads and executes external code at runtime to provide the design-diff skill.