rkit

L2 (Layer 2) reviewer for C/C++ language-specific idioms and safety patterns. Checks RAII, const correctness, smart pointers, security, concurrency, and sanitizer usage. Layer 2 of the 3-Layer code review architecture: L1 — Universal design quality (code-analyzer) L2 (this agent) — C/C++ language idioms L3 — Domain safety (safety-auditor, mcu-critical-analyzer) Triggers: C++ review, C review, cpp review, c-cpp review, C++ 리뷰, C 리뷰, C++レビュー, C++审查, revisión C++, revue C++, C++-Review, revisione C++ Do NOT use for: universal design quality (use code-analyzer), MISRA C rules (use safety-auditor), ISR/DMA safety (use mcu-critical-analyzer), or writing/modifying code (this agent is read-only).

You are a specialist in Claude Code CLI version analysis. Your mission is to

L1 (Layer 1) reviewer for universal, language-agnostic code design quality. Evaluates SOLID principles, complexity, DRY, naming, and anti-patterns. Produces a severity-classified review report (CRITICAL/HIGH/MEDIUM/LOW). Layer 1 of the 3-Layer code review architecture: L1 (this agent) — Universal design quality (domain-agnostic) L2 — Language-specific idioms (c-cpp-reviewer, csharp-reviewer, python-reviewer) L3 — Domain safety (safety-auditor, mcu-critical-analyzer, wpf-architect) Use proactively when user requests code review, quality check, or verification before PR / deployment. Triggers: code review, quality check, analyze, 코드 리뷰, 품질 검사, コードレビュー, 品質チェック, 代码审查, 质量检查, revisión de código, revue de code, Code-Review, revisione codice Do NOT use for: language-specific idioms (use c-cpp-reviewer / csharp-reviewer / python-reviewer), domain safety rules like MISRA/ISR/MVVM (use L3 agents), security-only scans (use security-architect), design document review (use design-validator), gap analysis (use gap-detector), or writing/modifying code (this agent is read-only).

L2 (Layer 2) reviewer for C# language-specific idioms and safety patterns. Checks async/await, nullable reference types, IDisposable, security, and sealed/IOptions patterns. Layer 2 of the 3-Layer code review architecture: L1 — Universal design quality (code-analyzer) L2 (this agent) — C# language idioms L3 — Domain safety (wpf-architect) Triggers: C# review, csharp review, dotnet review, C# 리뷰, C# レビュー, C#审查, revisión C#, revue C#, C#-Review, revisione C# Do NOT use for: universal design quality (use code-analyzer), WPF MVVM patterns (use wpf-architect), or writing/modifying code (this agent is read-only).

When spawned as an Agent Teams teammate, this agent operates as an independent

Lock architecture decisions from Design documents to prevent scope creep during Do phase. Auto-generates Mermaid diagrams for MCU memory maps, MPU software stacks, WPF MVVM structures. Triggers: arch-lock, architecture lock, lock design, 아키텍처 락, 설계 잠금, アーキテクチャロック, 設計ロック, 架构锁定, 设计锁定, bloqueo de arquitectura, verrouillage d'architecture, Architektursperre, blocco dell'architettura Do NOT use for: file-level freeze (use /freeze), automation control (use /control).

View audit logs, decision traces, and session history. Browse and search through rkit's audit trail for transparency and debugging. Use proactively when user wants to review AI decisions, check audit history, or investigate what actions were taken during a PDCA cycle. Triggers: audit, log, trace, history, decision, transparency, 감사, 로그, 추적, 이력, 결정, 투명성, 監査, ログ, 追跡, 履歴, 決定, 透明性, 审计, 日志, 追踪, 历史, 决策, 透明度, auditoria, registro, rastreo, historial, decision, transparencia, audit, journal, trace, historique, decision, transparence, Audit, Protokoll, Nachverfolgung, Verlauf, Entscheidung, Transparenz, audit, registro, traccia, cronologia, decisione, trasparenza Do NOT use for: modifying audit logs, PDCA phase execution, or code changes.

MCU/MPU/WPF domain-specific build and resource benchmarking. MCU: Flash/RAM usage via arm-none-eabi-size. MPU: rootfs/image size. WPF: build time and publish size. Triggers: benchmark, 벤치마크, ベンチマーク, 基准测试, benchmarking, Benchmark, riferimento

타겟 보드 디버깅 스킬. SSH 원격 테스트, 시리얼 로그 수집/분석, 디버그 리포트 생성. Triggers: 보드 디버깅, 보드 테스트, board debug, boot check, 부팅 확인

By-The-Way: 작업 중 개선 제안을 즉시 수집하고 관리하는 스킬. Triggers: /btw, /btw list, /btw analyze, /btw promote, /btw stats Keywords: btw, 개선, 제안, suggestion, improve, idea, feedback

11 hooks across 6 events

Concise, expert-level output for experienced embedded engineers

Beginner-friendly output with detailed explanations for embedded development

PDCA workflow + embedded expert level for advanced projects

PDCA workflow guide with phase status and domain-specific suggestions