This repository is for defensive AI evaluation and systems research. Please do not add operationally harmful content, credentials, private data, or internal project details.

For issues involving unsafe behavior, open a private report or contact yutoterashima4@gmail.com.