papers-skill

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

• Third-party content exposure detected (high risk: 0.85). Runtime path: search/detail/citations/arxiv call public APIs (Semantic Scholar api.semanticscholar.org and arXiv) and ingest returned free-text fields like title, abstract, tldr, and summary into the CLI output, which the agent then places into the LLM context—i.e., outsider-authored web content.