linkfox-lanjing-mercado-product-selection

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

• Third-party content exposure detected (high risk: 0.75). Outsider-authored free text can enter the agent’s LLM context via the runtime HTTP response body from the LinkFox gateway (POST /lingdong/call), where the script prints data/contentText-derived JSON to stdout (especially when --inline or response ≤ 8KB), and that printed text is then available to the agent for prompting.