Agent skill
a3criticalthinking
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/a3criticalthinking
SKILL.md
A3 Critical Thinking
When to Activate This Skill
- "Create an A3 for [problem]"
- "Help me think through [decision]"
- "Root cause analysis for [issue]"
- "What should take priority here?"
- "Is this safe to proceed?"
- "Evaluate tradeoffs for [options]"
- "5 Whys analysis"
- "Fishbone diagram for [defect]"
The Priority Hierarchy
Every decision must pass through this filter, in order:
┌─────────────────────────────────────────────────────┐
│ 1. SAFETY FIRST │
│ Will anyone be harmed? Stop everything else. │
│ • Employee safety │
│ • Customer safety (product in use) │
│ • Environmental safety │
│ • Community safety │
└─────────────────────────────────────────────────────┘
↓ Only if SAFE
┌─────────────────────────────────────────────────────┐
│ 2. CUSTOMER VALUE │
│ Does this create good products for customers? │
│ • Quality that meets/exceeds requirements │
│ • Reliability and durability │
│ • On-time delivery │
│ • Fitness for purpose │
└─────────────────────────────────────────────────────┘
↓ Only if QUALITY assured
┌─────────────────────────────────────────────────────┐
│ 3. SHAREHOLDER VALUE │
│ Now optimize for business results │
│ • Cost efficiency │
│ • Productivity │
│ • Return on investment │
│ • Growth and sustainability │
└─────────────────────────────────────────────────────┘
Critical Rule: Never sacrifice a higher priority for a lower one. A cost saving that compromises safety is NEVER acceptable. A delivery acceleration that reduces quality is NEVER acceptable.
The Decision Test
Before any significant decision, apply this test:
Question 1: Is it SAFE?
- Could this harm employees, customers, or the environment?
- Are all safety controls in place?
- Have we identified and mitigated risks?
- If NO: STOP. Address safety first.
Question 2: Does it serve the CUSTOMER?
- Will product quality be maintained or improved?
- Does this meet customer specifications?
- Will delivery commitments be met?
- If NO: STOP. Find an alternative that protects quality.
Question 3: Is it EFFICIENT?
- Only after safety and quality are assured, optimize for:
- Cost reduction
- Cycle time improvement
- Resource utilization
- Profitability
A3 Problem Solving Framework
The A3 is a single-page structured approach to problem solving:
┌─────────────────────────────────────────────────────────────────┐
│ TITLE: [Problem Name] DATE: │
│ OWNER: [Named Individual] REV: │
├─────────────────────────────────────────────────────────────────┤
│ 1. BACKGROUND/CONTEXT │ 2. CURRENT CONDITION │
│ │ │
│ Why is this problem important?│ What is actually happening? │
│ What triggered this A3? │ Data, facts, observations │
│ Business impact │ Process map of current state │
│ │ Quantify the gap │
├───────────────────────────────┼──────────────────────────────────┤
│ 3. TARGET CONDITION/GOAL │ 4. ROOT CAUSE ANALYSIS │
│ │ │
│ What should be happening? │ 5 Whys │
│ Specific, measurable target │ Fishbone/Ishikawa │
│ Timeline for achievement │ Data analysis │
│ │ Verified root cause(s) │
├───────────────────────────────┴──────────────────────────────────┤
│ 5. COUNTERMEASURES │
│ │
│ # │ Action │ Owner │ Due Date │ Status │
│ 1 │ │ │ │ │
│ 2 │ │ │ │ │
│ 3 │ │ │ │ │
├──────────────────────────────────────────────────────────────────┤
│ 6. IMPLEMENTATION PLAN │ 7. FOLLOW-UP/RESULTS │
│ │ │
│ Gantt or timeline │ Verification data │
│ Resources required │ Before/after comparison │
│ Risks and mitigation │ Lessons learned │
│ │ Horizontal deployment? │
└───────────────────────────────┴──────────────────────────────────┘
Root Cause Analysis Tools
5 Whys Method
Keep asking "Why?" until you reach the root cause (typically 5 levels):
Problem: Machine stopped producing
Why? → Fuse blew
Why? → Motor overheated
Why? → Bearing failed
Why? → Lubrication insufficient
Why? → No preventive maintenance schedule
ROOT CAUSE: Missing PM program for bearings
Rules:
- Each "Why" must be factual, not assumed
- Verify each level before proceeding
- May branch into multiple root causes
- Stop when you reach something you can control
Fishbone (Ishikawa) Diagram
Categorize potential causes:
Man Machine Material
\ | /
\ | /
\ | /
─────────[EFFECT]─────────
/ | \
/ | \
/ | \
Method Measurement Environment
Manufacturing Categories:
- Man/People: Training, skills, fatigue, following procedures
- Machine: Equipment condition, calibration, capability
- Material: Specifications, supplier quality, storage
- Method: Procedures, work instructions, sequence
- Measurement: Gages, accuracy, repeatability
- Environment: Temperature, humidity, cleanliness, lighting
Countermeasure Hierarchy
When developing solutions, prefer higher levels:
| Level | Type | Description | Example |
|---|---|---|---|
| 1 | Eliminate | Remove the possibility entirely | Design out the feature |
| 2 | Substitute | Replace with inherently safer/better | Different material |
| 3 | Engineer | Physical barriers or controls | Interlock, guard |
| 4 | Administrate | Procedures, training | Work instruction |
| 5 | PPE/Inspect | Last resort protection | Check, verify |
Rule: Never rely solely on administrative controls for safety-critical issues.
Quick Decision Framework
For rapid decisions under pressure:
┌─────────────────────────────────────────────┐
│ STOP AND ASK │
├─────────────────────────────────────────────┤
│ S - Safety: Is anyone at risk? │
│ T - Target: What are we trying to achieve? │
│ O - Options: What choices do we have? │
│ P - Priority: Safety → Quality → Cost │
└─────────────────────────────────────────────┘
If uncertain about safety: STOP PRODUCTION until verified safe.
Integration Points
With AutomotiveManufacturing Skill
- A3 links to PFMEA updates when new failure modes identified
- Countermeasures cascade to Work Instructions
- Control Plans updated based on A3 findings
With HoshinKanri Skill
- Red bowling chart items trigger A3
- A3 countermeasures become improvement priorities
- Completed A3s document breakthrough achievements
With Quality Systems (IATF 16949)
- A3 satisfies 10.2 Nonconformity and Corrective Action
- Links to 8D methodology for customer complaints
- Supports Management Review inputs
Templates Available
| Template | Purpose | Location |
|---|---|---|
| A3 Template | Standard problem solving | templates/a3-template.md |
| Quick A3 | Simplified one-pager | templates/quick-a3.md |
| 5 Whys | Root cause worksheet | templates/5-whys.md |
| Fishbone | Ishikawa diagram | templates/fishbone.md |
| Decision Matrix | Weighted option comparison | templates/decision-matrix.md |
| Priority Check | Safety-Quality-Cost verification | templates/priority-check.md |
Common Mistakes to Avoid
- Jumping to Solutions - Do root cause analysis first
- Blaming People - Look at systems and processes
- Stopping at Symptoms - Dig deeper with 5 Whys
- No Verification - Confirm countermeasures worked
- Ignoring the Hierarchy - Never shortcut Safety → Quality → Cost
- No Horizontal Deployment - Share learnings across similar processes
- Paper Exercise - A3 must drive real action
Key Principles
- Go See (Genchi Genbutsu) - Observe the actual condition yourself
- Facts Over Opinions - Base analysis on data and evidence
- Respect for People - Solutions should support workers, not blame them
- PDCA Cycle - Plan-Do-Check-Act is embedded in the A3
- One Problem, One Owner - Named individual accountability
- Visual Thinking - Use diagrams, charts, photos
- Priority Discipline - Safety → Customer → Shareholder, always
Extended Context
For detailed methodologies and advanced techniques:
read ~/.claude/skills/A3CriticalThinking/CLAUDE.md
For templates:
ls ~/.claude/skills/A3CriticalThinking/templates/
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?