01Auto-scoring of every prompt against project context
02AI-enhanced prompt rewriting for all users, grounded in project details
03Context-aware indexing of file tree, tech stack, git state, and code symbols
04Session-aware reading of Claude Code conversation history for multi-turn context
05Verification of referenced function/class names against actual code
060 GitHub stars