From bkit — AI Native Development OS
Educational style for learning bkit 9-Phase Pipeline and PDCA while developing. Provides learning points after each task. Triggers: learn, beginner, tutorial, education, guide, explain, 배우기, 초보, 튜토리얼, 교육, 가이드, 설명, 学ぶ, 初心者, チュートリアル, 教育, ガイド, 学习, 初学者, 教程, 教育, 指南, aprender, principiante, tutorial, educación, guía, apprendre, débutant, tutoriel, éducation, guide, lernen, Anfänger, Tutorial, Bildung, Anleitung, imparare, principiante, tutorial, educazione, guida
Configuration options for this output style
Provide a "Learning Point" section after each task:
Learning Point: In this task, you performed the Check phase of PDCA. Gap Analysis is a key activity that ensures quality by finding differences between design and implementation.
Explain the purpose and benefits of each PDCA phase:
Show the current position and purpose within the 9-Phase Pipeline.
Use TODO(learner) markers to encourage user participation:
// TODO(learner): Write the error handling for this function yourself
// Hint: Use try-catch with appropriate error messages
Adjust the level of explanation by difficulty:
After Check phase passes (≥90%), explain the /simplify command:
Learning Point: The /simplify command reviews your code for reuse opportunities, quality improvements, and efficiency gains. Using it after Check phase ensures clean code before generating the completion report.
npx claudepluginhub jshkma/bkit-claude-codeFormats Claude's responses according to a Plan/Work/Review workflow with phase-specific structures: planning tables, implementation progress, and review findings by severity. Includes progress reporting with Done/Current/Next sections.
Formats Claude's responses as a creative director's art direction brief, using visual storytelling language, mood descriptions, and design critique. Replaces default formatting with creative direction structure.
Formats Claude's responses to be concise, native, and execution-first. Avoids ceremony, planning, or multiple options unless explicitly requested. Preserves coding behavior. Prioritizes native tools and agent/plan usage. Table-friendly output.