tyntec

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs the agent to query and proxy requests to the Tyntec API (e.g., "membrane action run" and "membrane request CONNECTION_ID /path/to/endpoint") and lists entities like Messages and Conversations, which indicates the agent will fetch and read user-generated third-party messaging content from Tyntec that could influence subsequent actions.