guided-testimonial-video

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly asks the user for a marketing site URL (STEP 1) and then instructs the agent to fetch and extract brand signals from that live public site (STEP 3), which is untrusted third-party content that the agent must read and use to drive creative decisions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches a user-provided marketing site URL at runtime ("Are we inside your brand's repository, or would you like to provide a URL (e.g. your marketing site) so I can research your brand online?") and then directly uses the fetched site content to build the brand profile and drive the creative-spec prompts, so an arbitrary external URL supplied at runtime can control the agent's instructions.