Skill

mcu-critical-analysis

MCU 동시성/시스템 크리티컬 이슈 통합 분석 스킬. ConSynergy 4-Stage 파이프라인으로 ISR/DMA/동시성/스택 이슈를 분석하고 심각도×신뢰도 분류 리포트를 생성합니다. Use proactively when user mentions concurrency issues, race conditions, interrupt safety, DMA conflicts, or asks for critical code review. Triggers: mcu-critical-analysis, critical analysis, 크리티컬 분석, 동시성 분석, 동시성 감사, concurrency, concurrency audit, race condition, 레이스 컨디션, data race, deadlock, 데드락, ISR 분석, interrupt analysis, DMA 분석, 스택 분석, クリティカル分析, 并发分析, análisis de concurrencia, Konkurrenzanalyse, analisi di concorrenza Do NOT use for: 일반 코드 리뷰 (code-review 사용), MISRA 검증만 필요한 경우 (misra-c 사용), 아키텍처 설계 (fw-architect 에이전트 사용).

Install

npx claudepluginhub solitasroh/rkit --plugin rkit

Tool Access

This skill is limited to using the following tools:

ReadWriteEditGlobGrepBashTask

Preview

> Detects concurrency bugs, race conditions, and system-critical issues in MCU firmware

SKILL.md

Similar Skills

using-git-worktrees

Creates isolated Git worktrees for feature branches with prioritized directory selection, gitignore safety checks, auto project setup for Node/Python/Rust/Go, and baseline verification.

superpowers

169.2k

subagent-driven-development

3 files

Executes implementation plans in current session by dispatching fresh subagents per independent task, with two-stage reviews: spec compliance then code quality.

superpowers

169.2k

dispatching-parallel-agents

Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.

superpowers

169.2k

Stats

Stars2

Forks0

Last CommitApr 3, 2026

Actions

View Source View Plugin View on GitHub View README

MCU Critical Analysis Skill

Detects concurrency bugs, race conditions, and system-critical issues in MCU firmware using a 2-Layer architecture: JS extraction modules + LLM reasoning agent.

Overview

This skill orchestrates:

Layer 1 (JS Extraction): Runs 4 extraction modules to collect structured data
Layer 2 (LLM Reasoning): Invokes mcu-critical-analyzer agent for semantic analysis

Workflow

Step 1: Determine Source Directory

If $ARGUMENTS is provided, use it as srcDir
Otherwise, search current directory for .c/.h files
Find .ioc file if present (optional, enhances NVIC/DMA analysis)
Find .map file if present (optional, enhances stack analysis)
Find build directory with .su files if present (optional)

Step 2: Run Layer 1 Extraction

Execute extraction functions from lib/mcu/ modules using Bash:

node -e "
const mcu = require('./lib/mcu');
const srcDir = '{srcDir}';
const iocPath = '{iocPath}'; // or null

// ISR extraction
const handlers = mcu.extractISRHandlers(srcDir);
const nvicConfig = mcu.extractNVICConfig(iocPath || null, srcDir);
const callGraph = mcu.extractISRCallGraph(srcDir, handlers);
const isrGlobalAccess = mcu.extractISRGlobalAccess(srcDir, handlers);

// DMA extraction
const dmaConfig = mcu.extractDMAConfig(iocPath || null, srcDir);
const dmaBuffers = mcu.extractDMABuffers(srcDir);

// Concurrency extraction
const globalVars = mcu.extractGlobalVariables(srcDir);
const syncPrimitives = mcu.extractSyncPrimitives(srcDir);
const switchPoints = mcu.extractContextSwitchPoints(srcDir);
const atomicOps = mcu.extractAtomicOps(srcDir);

// Tool bridge (optional)
const cppcheck = mcu.runCppcheckThreadsafety(srcDir);
const stackUsage = mcu.extractStackUsage('{buildDir}', '{mapPath}');

// Output summary
console.log(JSON.stringify({
  handlers: handlers.length,
  nvicConfig: nvicConfig.length,
  dmaChannels: dmaConfig.length,
  dmaBuffers: dmaBuffers.length,
  globalVars: globalVars.length,
  syncPrimitives: syncPrimitives.length,
  switchPoints: switchPoints.length,
  atomicOps: atomicOps.length,
  cppcheckAvailable: cppcheck.available,
  cppcheckFindings: cppcheck.findings.length,
  stackAvailable: stackUsage.available,
  stackEntries: stackUsage.entries.length,
}, null, 2));
"

Step 3: Invoke mcu-critical-analyzer Agent

Pass extraction results to the agent. The agent will:

Read refs/mcu/concurrency-patterns.md and custom-scheduler-guide.md
Execute ConSynergy 4-Stage Pipeline (Identify → Slice → Reason → Verdict)
Read actual source code at flagged locations for verification
Generate classified issue report

Step 4: Output Report

Save report to docs/03-analysis/mcu-critical/{project}-{YYYYMMDD}.md
Display summary in response:

MCU Critical Analysis Complete
─────────────────────────────
Critical: N issues (M high confidence)
Warning:  N issues
Info:     N issues
─────────────────────────────
Report: docs/03-analysis/mcu-critical/...
Next: Review critical issues and apply fixes

Usage Examples

# Analyze current project
/mcu-critical-analysis

# Analyze specific source directory
/mcu-critical-analysis src/

# Analyze after making changes
/mcu-critical-analysis firmware/Core/Src

Integration with PDCA

This skill integrates into the PDCA cycle:

Typically used during Do phase (implementation review)
Results feed into Check phase (/pdca analyze)
Critical issues should be fixed before /pdca analyze for better match rate