marketing-sostac

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

• Third-party content exposure detected (high risk: 1.00). The skill's Auto-Discovery and Phase workflows (see ./references/auto-discovery.md and the SKILL.md "Pre-Phase 0: Auto-Discovery" and "Research Sequence") explicitly instruct the agent to fetch and scrape open/public third‑party sources (brand websites, Google SERP, competitor sites, Reddit, Quora, G2/Trustpilot, Meta Ad Library, SimilarWeb, LinkedIn, etc.) and to read and interpret that untrusted user-generated/public content to drive research, recommendations, and subsequent automated actions, which meets the conditions for indirect prompt injection risk.