efecto-web-design

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to fetch and import third-party CSS (e.g., "read their CSS file" / the user's globals.css) and to call the built-in image search (Lummi) returning external image URLs — both are untrusted external content the agent must read and apply (e.g., via set_theme/css and using image URLs), which can materially change theme tokens and downstream tool actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires installing and running the Efecto MCP server with "npx -y @efectoapp/mcp", which fetches and executes remote code at runtime and is a required dependency for the skill's operation.