paper-analyzer

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md workflow (Step 1) explicitly instructs adding arbitrary arXiv HTML/abstract URLs into NotebookLM (e.g., notebooklm source add "https://arxiv.org/html/") and then using notebooklm ask to read and interpret those papers, so the agent ingests untrusted public third-party content that can materially influence its analysis and actions.