dark-funnel

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

• Third-party content exposure detected (high risk: 0.90). The skill's Process step 2 and Agents Used explicitly state it collects and normalizes data from public community platforms (e.g., "Reddit and community mention frequency and sentiment"), podcasts, AI chatbot citation data, and other third-party sources, meaning the agent ingests untrusted user-generated content that can influence analysis and subsequent actions.