counter-narrative

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

• Third-party content exposure detected (high risk: 1.00). The skill's required "Evidence" input explicitly asks for URLs and artifacts (e.g., competitor landing pages, social posts, press releases) and step 2 directs the agent to examine that evidence, meaning the agent will fetch and interpret untrusted public web content as part of its workflow.