kermt-continue-pretrain

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

• Third-party content exposure detected (medium risk: 0.30). The workflow ingests outsider-authored free text via the user-supplied --csv <path> which is mounted into the container and consumed by python agent/scripts/check_data.py and python agent/scripts/prepare_data.py during the “Validate the data” and “Prepare the data” steps.