google-adk-python

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly integrates a google_search tool and example agents (e.g., "search_assistant" and "Researcher") that perform web searches and summarization, causing the agent to fetch and read arbitrary public web content (including public GitHub/raw URLs) which is untrusted and could enable indirect prompt injection.