tawkto

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to use Membrane actions and the "proxy requests" flow to fetch and interact with Tawk.to data (e.g., Visitor chats and Knowledge Base Articles) — user-generated content from a third-party API that the agent will read and act on, creating a clear avenue for indirect prompt injection.