librarian

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

• Third-party content exposure detected (medium risk: 0.30). For /librarian from gmail and /librarian from web, the workflow calls Gmail’s search_threads/get_thread or web extractors (web_fetch_exa, tavily_extract) on user-selected items and ingests their message bodies/page markdown into vault notes (i.e., outsider-authored free text can be directly read by the LLM during extraction and synthesis).