exa-context

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly runs exa-ai search and webset commands (e.g., "exa-ai search 'AI news'" and the webset-create/webset-search-create examples) which fetch and ingest public web content for context, so the agent will read untrusted third-party (web/social/forum/news) content as part of its workflow.