scientific-literature-search

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md explicitly instructs the agent to use tools that fetch and extract content from public web sources and arbitrary URLs (advanced_web_search_claude, search_google, extract_url_content, extract_pdf_content, fetch_supplementary_info_from_doi — e.g., protocols.io, supplier sites, arbitrary webpages), and to read and act on that content as part of its search and extraction workflow, exposing it to untrusted third-party content that could carry indirect prompt injections.