01Three-level progressive disclosure architecture for context efficiency
02Mandatory /reference/ folder structure for deferred resource loading
03Automated validation script to verify skill structure and compliance
041 GitHub stars
05Pre-built SKILL.md templates and step-by-step creation processes
06Strict token optimization budgets (Metadata ~100, Body <2k tokens)