seo-strategy

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly instructs the agent to gather and analyze open/public web content (e.g., "Google Autocomplete," "People Also Ask," and "Reddit/niche forums" in Step 2 and repeated manual SERP/competitor content analysis), which are untrusted, user-generated public sources the agent would read/interpret as part of its workflow and could allow indirect prompt injection.