content-research-writer

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

• Third-party content exposure detected (high risk: 0.72). The skill’s required workflow includes “Research … and add citations” (SKILL.md:151-156), which—at runtime—would necessarily involve fetching/extracting external sources (public web content or other outsider-authored text) that the agent can ingest into LLM context.