vue-layout

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

• Third-party content exposure detected (medium risk: 0.30). Skill.md requires ingesting outsider-authored free text via Step 1 by accepting a user-provided Figma URL/node-id and then using the Framelink MCP to read Figma node structures/styles (and screenshots) in Step 2/5, where Figma content can include arbitrary text inserted by the submitter.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill requires the user-provided Figma URL (https://figma.com/design/{fileKey}/{fileName}?node-id={nodeId}) and explicitly calls Framelink MCP functions at runtime to fetch Figma data/images that are injected into the agent's processing to drive code generation, so this external URL is a runtime dependency that directly influences prompts/output.