Literature Review

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly instructs the agent to use WebSearch/Firecrawl to fetch and read content from open public sources (e.g., Google Scholar, PubMed, arXiv, JSTOR), so it ingests third-party web content that could contain untrusted or user-generated material enabling indirect prompt injection.