tavily

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

• Third-party content exposure detected (high risk: 0.90). The Tavily skill issues live web searches to https://api.tavily.com/search and returns results plus optional raw page content (snippets/HTML) from arbitrary public websites — including social sites referenced in examples — which the agent is expected to read and summarize, exposing it to untrusted third‑party content that could enable indirect prompt injection.