Effect-based safety gate for AI coding agents' shell commands (OpenCode, Antigravity): fast structural rules, then TypeSafe's Jev or a chat model judges what a command does. Certified with Jev at zero dangerous commands allowed.
command classification using Keras
Stop clicking Allow on every safe command
Command classification system using Logistic Regression and MLflow. Trains, tracks, and registers models with version control, metrics logging, and production deployment. Built with scikit-learn and Python for ML lifecycle management.
To associate your repository with the command-classification topic, visit your repo's landing page and select "manage topics."