deep-research

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs the agent to use web search/consensus MCP and to fetch public sources (company sites, news, real-user reviews, primary documents) and requires working links in the "Plan the research" and "Rules" sections, so the skill expects reading and acting on untrusted third‑party web content.