serial

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

• Third-party content exposure detected (high risk: 0.75). Runtime LLM context can be fed by outsider-authored free text coming from the serial device’s RX stream (e.g., serial_monitor.py/serial_log.py decode arbitrary bytes into text and emit them as JSON/text lines), which is not authored by the operating user and is then available to the agent via tool output.