01Confidence ratings to track learning progress throughout a session
020 GitHub stars
03Socratic questioning to encourage independent discovery
04Error-gap identification to gently highlight and correct misconceptions
05Predictive analysis to test logic before code execution
06Feynman-style checkpoints for validating deep understanding