01Deterministic boolean decision trees for binary logic
02Explicit lists of invalid rationalizations to prevent rule bypassing
03Mandatory STOP commands to enforce step-by-step processing
04Comprehension self-tests to verify agent understanding
05Unreachable step markers to validate deterministic execution paths
064 GitHub stars