academic-scientific-research

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

• Third-party content exposure detected (high risk: 0.85). This skill’s required workflow is to “use search” and “extract from the strongest selected pages,” which at runtime would ingest outsider-authored free text from public web/scholarly pages (e.g., articles, PubMed/journal pages) into the LLM context via the companion search/extract steps.