social-science-paper-writing

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

• Third-party content exposure detected (high risk: 0.75). The skill’s literature_search_to_review mode can ingest outsider-authored free text from runtime-fetched public web content (e.g., CNKI/Google Scholar result pages and any retrieved abstracts/full texts) into the agent’s LLM context while screening/reading sources.