bmad-agent-marketing-psychology

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md and Shared Patterns explicitly require auditing a "Live Website URL" and instruct using agent-browser or WebFetch/WebSearch to fetch and inspect public webpages as the working source of truth (e.g., "Live URL → audit the live experience first" and the "agent-browser Setup"), so it will read untrusted, user-provided public web content and use that content to drive diagnostic decisions and recommended actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's agent-browser setup explicitly instructs installing and running remote code at runtime from the GitHub URL https://github.com/vercel-labs/agent-browser (via "npx skills add https://github.com/vercel-labs/agent-browser" / "npm install -g agent-browser"), which fetches and executes external code the skill may rely on for browser-based research—meeting the criteria for a runtime external dependency that can execute code.