transformation-workflow
Practical application guide for HUMMBL's 6 transformations (Perspective, Inversion, Composition, Decomposition, Recursion, Meta-Systems). Includes when to use each transformation, combination patterns, analysis templates, output formats, real-world examples, and common pitfalls. Essential for applying mental models effectively in problem-solving and analysis.
Best use case
transformation-workflow is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Practical application guide for HUMMBL's 6 transformations (Perspective, Inversion, Composition, Decomposition, Recursion, Meta-Systems). Includes when to use each transformation, combination patterns, analysis templates, output formats, real-world examples, and common pitfalls. Essential for applying mental models effectively in problem-solving and analysis.
Teams using transformation-workflow should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/transformation-workflow/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How transformation-workflow Compares
| Feature / Agent | transformation-workflow | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Practical application guide for HUMMBL's 6 transformations (Perspective, Inversion, Composition, Decomposition, Recursion, Meta-Systems). Includes when to use each transformation, combination patterns, analysis templates, output formats, real-world examples, and common pitfalls. Essential for applying mental models effectively in problem-solving and analysis.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Transformation Workflow Skill
Practical guide for applying HUMMBL's 6 transformations to real-world problems. Provides step-by-step workflows, combination patterns, templates, and examples for effective mental model usage.
## Overview
The 6 HUMMBL transformations represent different cognitive operations:
1. **Perspective (P):** Frame and name what is
2. **Inversion (IN):** Reverse assumptions
3. **Composition (CO):** Combine parts into wholes
4. **Decomposition (DE):** Break wholes into components
5. **Recursion (RE):** Iterate, feedback, self-reference
6. **Meta-Systems (SY):** Coordinate systems-of-systems
## When to Use Each Transformation
### Perspective (P) - Use When:
**Problem Indicators:**
- ✅ Problem statement unclear or ambiguous
- ✅ Stakeholders have conflicting views
- ✅ Need to understand different viewpoints
- ✅ Framing feels wrong or limiting
- ✅ Context not fully understood
**Trigger Questions:**
- "How do different people see this?"
- "What am I missing in how I frame this?"
- "Whose perspective matters here?"
- "What context am I ignoring?"
**Best For:**
- Requirements gathering
- Stakeholder analysis
- Problem definition
- User research
- Strategic framing
### Inversion (IN) - Use When:
**Problem Indicators:**
- ✅ Stuck with conventional thinking
- ✅ Need fresh perspective
- ✅ Want to avoid failure
- ✅ Looking for non-obvious solutions
- ✅ Need to challenge assumptions
**Trigger Questions:**
- "What's the opposite approach?"
- "What if this fails - why?"
- "What should we NOT do?"
- "What assumptions can we reverse?"
**Best For:**
- Brainstorming
- Risk analysis
- Innovation
- Assumption testing
- Creativity boost
### Composition (CO) - Use When:
**Problem Indicators:**
- ✅ Have multiple components to integrate
- ✅ Need to build cohesive solution
- ✅ Want synergies between parts
- ✅ Creating system from pieces
- ✅ Assembling team/resources
**Trigger Questions:**
- "How do these parts work together?"
- "What synergies exist?"
- "How to integrate this?"
- "What's the whole picture?"
**Best For:**
- Solution design
- System architecture
- Team formation
- Strategy synthesis
- Product development
### Decomposition (DE) - Use When:
**Problem Indicators:**
- ✅ System too complex to understand
- ✅ Need to find root cause
- ✅ Looking for bottlenecks
- ✅ Want to prioritize efforts
- ✅ Debugging or troubleshooting
**Trigger Questions:**
- "What are the parts?"
- "Why is this happening?"
- "Where's the constraint?"
- "What's essential vs nice-to-have?"
**Best For:**
- Problem diagnosis
- System analysis
- Prioritization
- Root cause analysis
- Debugging
### Recursion (RE) - Use When:
**Problem Indicators:**
- ✅ Dealing with feedback loops
- ✅ Iterative process needed
- ✅ Self-reinforcing dynamics present
- ✅ Need progressive improvement
- ✅ Growth/decline accelerating
**Trigger Questions:**
- "What's feeding back into itself?"
- "How do we iterate?"
- "What cycles exist here?"
- "What's the second-order effect?"
**Best For:**
- Growth strategy
- Process improvement
- System dynamics
- Iterative development
- Feedback management
### Meta-Systems (SY) - Use When:
**Problem Indicators:**
- ✅ Strategic decision needed
- ✅ Multiple systems interacting
- ✅ Long-term consequences matter
- ✅ Systemic intervention needed
- ✅ Choosing which model to use
**Trigger Questions:**
- "What's the systems view?"
- "What are second/third-order effects?"
- "Where's the leverage point?"
- "Which mental model applies?"
**Best For:**
- Strategic planning
- System design
- Leverage point identification
- Model selection
- Long-term thinking
## Transformation Workflows
### Workflow 1: Perspective Analysis
**Input:** Problem statement, context
**Steps:**
1. **State the problem** (1 sentence)
2. **List stakeholders** (P2: Stakeholder Mapping)
- Who is affected?
- Who has power?
- Who has information?
3. **Apply multiple lenses** (P4: Lens Shifting)
- Technical lens
- Business lens
- User lens
- Ethical lens
4. **Identify first principles** (P1)
- What must be true?
- What are non-negotiables?
- What are fundamental constraints?
5. **Document context** (P8: Context Awareness)
- Time constraints
- Resource constraints
- Political/cultural factors
**Output Format:**
```markdown
## Perspective Analysis
**Problem:** [1-sentence problem statement]
**Stakeholders:**
- [Stakeholder 1]: [Their perspective/interest]
- [Stakeholder 2]: [Their perspective/interest]
- [Stakeholder 3]: [Their perspective/interest]
**Multiple Lenses:**
- **Technical:** [Technical view]
- **Business:** [Business view]
- **User:** [User view]
- **Ethical:** [Ethical considerations]
**First Principles:**
1. [Fundamental truth 1]
2. [Fundamental truth 2]
3. [Fundamental truth 3]
**Context:**
- **Time:** [Timeline factors]
- **Resources:** [Resource constraints]
- **Environment:** [External factors]
**Insights:**
- [Key insight 1]
- [Key insight 2]
```
**Example:** Software architecture decision
- **Problem:** Choose between microservices vs monolith
- **Stakeholders:** Engineering (prefers interesting tech), Product (wants speed), Operations (wants stability)
- **Lenses:** Technical (complexity trade-offs), Business (cost/time), User (performance)
- **First Principles:** Team size matters more than technology
- **Output:** Decision framework based on team constraints, not tech fashion
### Workflow 2: Inversion Analysis
**Input:** Problem, current approach
**Steps:**
1. **State current approach**
2. **Apply inversion** (IN1)
- What if we did the opposite?
- What would the inverse solution look like?
3. **Run premortem** (IN8)
- Assume total failure in 6 months
- Why did it fail?
- What went wrong?
4. **Apply via negativa** (IN3)
- What should we STOP doing?
- What to remove, not add?
5. **Seek disconfirmation** (IN15)
- What evidence contradicts our plan?
- Who disagrees and why?
**Output Format:**
```markdown
## Inversion Analysis
**Current Approach:** [Description]
**Inverted Approach:**
- Instead of [X], what if we [opposite of X]?
- Result: [Insights from inversion]
**Premortem (Assume Failure):**
- **Failure Scenario:** [What failed]
- **Root Cause:** [Why it failed]
- **Warning Signs:** [Early indicators we missed]
**Via Negativa (What to STOP):**
- Stop: [Thing 1]
- Stop: [Thing 2]
- Stop: [Thing 3]
**Disconfirming Evidence:**
- [Evidence against our approach]
- [Counterargument]
- [Risk we're underestimating]
**Revised Approach:**
- [Improvements based on inversion]
```
**Example:** Product launch strategy
- **Current:** Big launch event, lots of marketing
- **Inversion:** What if we did quiet launch to small group?
- **Premortem:** Event flops because nobody cares, spent budget wrong
- **Via Negativa:** Stop assuming launch is most important thing
- **Output:** Phased launch, test with early adopters first
### Workflow 3: Composition Strategy
**Input:** Components, requirements
**Steps:**
1. **List all components**
2. **Identify synergies** (CO1)
- Where do parts enhance each other?
- What emergent properties arise?
3. **Design synthesis** (CO4)
- How to merge into coherent whole?
- What's the unifying concept?
4. **Plan orchestration** (CO19)
- How to coordinate components?
- What's the execution sequence?
5. **Create holistic integration** (CO20)
- Complete unified system
- No loose ends
**Output Format:**
```markdown
## Composition Strategy
**Components:**
1. [Component 1] - [Purpose]
2. [Component 2] - [Purpose]
3. [Component 3] - [Purpose]
**Synergies:**
- [Comp A] + [Comp B] = [Synergy]
- [Comp B] + [Comp C] = [Synergy]
**Synthesis Design:**
- **Unifying Concept:** [Central idea that ties everything]
- **Integration Points:** [Where components connect]
- **Emergent Properties:** [New capabilities from combination]
**Orchestration Plan:**
1. [Phase 1]: [Components + actions]
2. [Phase 2]: [Components + actions]
3. [Phase 3]: [Components + actions]
**Holistic Integration:**
- [How all pieces form complete system]
- [Quality properties of whole]
```
**Example:** Building product ecosystem
- **Components:** Core product, API, marketplace, analytics
- **Synergies:** API enables marketplace, marketplace drives analytics, analytics improves product
- **Synthesis:** Platform strategy
- **Output:** Integrated ecosystem with network effects
### Workflow 4: Decomposition Analysis
**Input:** Complex system or problem
**Steps:**
1. **Define the whole**
2. **Find root cause** (DE1)
- 5 Whys technique
- Causal chain analysis
3. **Apply divide & conquer** (DE2)
- Break into logical subsystems
- Identify interfaces
4. **Identify bottleneck** (DE6)
- Theory of Constraints
- What's the limiting factor?
5. **Pareto analysis** (DE7)
- What's the vital 20%?
- Where to focus effort?
**Output Format:**
```markdown
## Decomposition Analysis
**System:** [Description of whole]
**Root Cause Analysis:**
- Why? [Reason 1]
- Why? [Reason 2]
- Why? [Reason 3]
- Why? [Reason 4]
- Why? [ROOT CAUSE]
**Component Breakdown:**
├── [Component A]
│ ├── [Subcomponent A1]
│ └── [Subcomponent A2]
├── [Component B]
│ ├── [Subcomponent B1]
│ └── [Subcomponent B2]
└── [Component C]
**Bottleneck:**
- **Constraint:** [Limiting factor]
- **Impact:** [How it limits system]
- **Intervention:** [How to address]
**Pareto (80/20):**
- **Vital Few (20%):**
- [Critical element 1]
- [Critical element 2]
- **Trivial Many (80%):**
- [Less critical elements]
**Action Plan:**
1. [Address root cause]
2. [Remove bottleneck]
3. [Focus on vital 20%]
```
**Example:** Website performance issues
- **Root Cause:** Inefficient database queries (not server capacity)
- **Breakdown:** Frontend, API, Database, Cache, CDN
- **Bottleneck:** Database query on user table
- **Pareto:** 3 queries cause 80% of slow responses
- **Output:** Optimize those 3 queries first
### Workflow 5: Recursion Analysis
**Input:** System with dynamics over time
**Steps:**
1. **Map feedback loops** (RE1)
- Positive (reinforcing)
- Negative (balancing)
2. **Identify virtuous cycles** (RE7)
- What creates growth?
- How to amplify?
3. **Identify vicious cycles** (RE8)
- What creates decline?
- How to break?
4. **Design iteration** (RE2)
- How to improve progressively?
- What's the learning loop?
5. **Analyze second-order** (RE19)
- Effects of effects
- Compound dynamics
**Output Format:**
```markdown
## Recursion Analysis
**System Dynamics:**
**Feedback Loops:**
- ➕ **Virtuous Cycle:** [A] → [B] → [C] → [More A]
- ➖ **Vicious Cycle:** [X] → [Y] → [Z] → [More X]
- ⚖️ **Balancing Loop:** [M] → [N] → [Less M]
**Virtuous Cycles (Amplify These):**
1. [Positive cycle 1]
- Trigger: [What starts it]
- Amplify: [How to strengthen]
2. [Positive cycle 2]
**Vicious Cycles (Break These):**
1. [Negative cycle 1]
- Cause: [What perpetuates it]
- Intervention: [How to break]
2. [Negative cycle 2]
**Iterative Improvement:**
- **Version 1:** [Initial state]
- **Learn:** [What to measure]
- **Improve:** [What to adjust]
- **Repeat:** [Cycle time]
**Second-Order Effects:**
- First-order: [Direct effect]
- Second-order: [Effect of effect]
- Third-order: [Effect of effect of effect]
**Leverage Points:**
- [Where small change creates big impact]
```
**Example:** SaaS growth
- **Virtuous Cycle:** Good product → Happy users → Referrals → More users → More feedback → Better product
- **Vicious Cycle:** Bugs → Bad reviews → Fewer signups → Less revenue → Less engineering → More bugs
- **Iteration:** Weekly releases, measure NPS, improve top complaint
- **Output:** Strategy to amplify virtuous, break vicious cycles
### Workflow 6: Meta-Systems Strategy
**Input:** Strategic question or complex system
**Steps:**
1. **Apply systems thinking** (SY1)
- See whole system
- Identify interconnections
2. **Second-order thinking** (SY2)
- Consequences of consequences
- Nth-order effects
3. **Find leverage points** (SY4)
- Where to intervene?
- High-impact, low-effort
4. **Anticipate unintended consequences** (SY5)
- What could go wrong?
- Side effects?
5. **Model selection** (SY19)
- Which other models apply?
- What's the right analytical approach?
**Output Format:**
```markdown
## Meta-Systems Strategy
**Strategic Question:** [Question]
**Systems View:**
┌─────────────────────────────────┐
│ [System Component 1] │
│ ↓ ↑ │
│ [System Component 2] │
│ ↓ ↑ │
│ [System Component 3] │
└─────────────────────────────────┘
**Interconnections:**
- [A] affects [B] via [mechanism]
- [B] affects [C] via [mechanism]
- [C] feeds back to [A] via [mechanism]
**Second-Order Analysis:**
| Action | 1st Order | 2nd Order | 3rd Order |
|--------|-----------|-----------|-----------|
| [Action 1] | [Direct effect] | [Effect of effect] | [Further effect] |
| [Action 2] | [Direct effect] | [Effect of effect] | [Further effect] |
**Leverage Points** (Highest to Lowest Impact):
1. **[Point 1]:** [Why high leverage]
2. **[Point 2]:** [Why medium leverage]
3. **[Point 3]:** [Why low leverage]
**Unintended Consequences:**
- Risk: [Potential negative outcome]
- Mitigation: [How to prevent]
**Model Selection:**
- Primary: [Model code + name]
- Secondary: [Model code + name]
- Why: [Justification]
**Recommended Strategy:**
- [Strategic approach based on analysis]
```
**Example:** Market expansion decision
- **Systems View:** Current market, new market, competitors, resources
- **Second-Order:** Enter new market → Spread resources thin → Lose focus in current market → Competitors gain ground
- **Leverage:** Instead of new market, deepen penetration in current (10x ROI)
- **Output:** Stay focused strategy, not expansion
## Combination Patterns
### Pattern 1: P → DE → CO (Understand → Analyze → Build)
**Use Case:** Building new solution
**Steps:**
1. **Perspective:** Understand problem from multiple angles
2. **Decomposition:** Break down into components
3. **Composition:** Integrate into solution
**Example:** Designing new feature
- P: Stakeholder needs (users want X, business wants Y)
- DE: Break into sub-features, identify dependencies
- CO: Integrate into cohesive feature with good UX
### Pattern 2: P → IN → SY (Frame → Challenge → Strategy)
**Use Case:** Strategic decision
**Steps:**
1. **Perspective:** Frame the situation
2. **Inversion:** Challenge assumptions
3. **Meta-Systems:** Strategic synthesis
**Example:** Business model pivot
- P: Current model's perspective, customer viewpoint
- IN: What if opposite? What to stop?
- SY: Strategic choice based on systems thinking
### Pattern 3: DE → IN → CO (Analyze → Invert → Rebuild)
**Use Case:** Innovation/redesign
**Steps:**
1. **Decomposition:** Understand current system
2. **Inversion:** Challenge how it works
3. **Composition:** Build new solution
**Example:** Process improvement
- DE: Map current process, find bottleneck
- IN: What if we removed steps? Did opposite?
- CO: Redesigned process
### Pattern 4: All 6 in Sequence (Complete Analysis)
**Use Case:** Major strategic initiative
**Steps:**
1. **P:** Frame problem
2. **IN:** Challenge assumptions
3. **DE:** Analyze components
4. **CO:** Build solution
5. **RE:** Plan iteration
6. **SY:** Strategic integration
**Example:** Company transformation
- Use all 6 transformations systematically
- Comprehensive, robust analysis
- Takes longer but minimizes blind spots
### Pattern 5: RE wrapping any other (Iterative Application)
**Use Case:** Continuous improvement
**Structure:** RE(P/IN/CO/DE/SY)
**Example:** Product development
- Week 1: P (understand users)
- Week 2: DE (analyze feedback)
- Week 3: CO (build improvements)
- Week 4: RE (iterate based on results)
- Repeat
## Common Pitfalls & Solutions
### Pitfall 1: Using Wrong Transformation
**Error:** Applying Decomposition when need Perspective
**Symptom:** Breaking down problem doesn't help because problem not understood
**Solution:** Start with P (frame first), then DE (analyze)
### Pitfall 2: Skipping Inversion
**Error:** Going straight to solution without challenging assumptions
**Symptom:** Conventional thinking, missing creative options
**Solution:** Always apply IN before finalizing approach
### Pitfall 3: Decomposition Without Recomposition
**Error:** Breaking things down but never synthesizing
**Symptom:** Analysis paralysis, no actionable solution
**Solution:** DE must be followed by CO (analyze then build)
### Pitfall 4: Ignoring Feedback Loops
**Error:** Linear thinking in dynamic system
**Symptom:** Interventions don't work as expected
**Solution:** Apply RE to understand dynamics
### Pitfall 5: Local Optimization
**Error:** Optimizing parts without seeing whole
**Symptom:** Suboptimization, missing systemic issues
**Solution:** Use SY (systems view) before optimizing
### Pitfall 6: Single-Model Thinking
**Error:** Using only one model/transformation
**Symptom:** One-dimensional analysis, blind spots
**Solution:** Combine multiple transformations (patterns above)
### Pitfall 7: Overcomplication
**Error:** Applying all 6 when 2 would suffice
**Symptom:** Slow progress, diminishing returns
**Solution:** Start simple (1-2 transformations), add if needed
## Transformation Selection Flowchart
```
START: What's your primary need?
├─ "Understand the problem"
│ → Use PERSPECTIVE (P)
│ → Then consider: DE (analyze) or IN (challenge)
├─ "Stuck or need creativity"
│ → Use INVERSION (IN)
│ → Then consider: P (reframe) or CO (rebuild)
├─ "Build/integrate solution"
│ → Use COMPOSITION (CO)
│ → Likely needed: DE first (analyze parts)
├─ "Analyze complex system"
│ → Use DECOMPOSITION (DE)
│ → Then consider: CO (reintegrate) or SY (systems view)
├─ "Handle dynamics/feedback"
│ → Use RECURSION (RE)
│ → Then consider: SY (systemic) or DE (analyze loops)
└─ "Strategic/systemic decision"
→ Use META-SYSTEMS (SY)
→ Then consider: P (perspectives) + IN (challenge)
```
## Quick Templates
### 5-Minute Quick Analysis
1. **P:** Who are stakeholders? (30 sec)
2. **IN:** What's the opposite? (30 sec)
3. **DE:** What's the bottleneck? (1 min)
4. **CO:** How to integrate? (1 min)
5. **RE:** What's the feedback? (1 min)
6. **SY:** What's the leverage? (1 min)
### One-Page Strategy
**Problem:** [1 sentence]
**Perspective:** [Key stakeholders, key lens]
**Inversion:** [What NOT to do]
**Decomposition:** [Critical components]
**Composition:** [How they integrate]
**Recursion:** [Key feedback loop]
**Systems:** [Leverage point]
**Action:** [Next step]
## Resources
- **HUMMBL Framework Skill:** Complete model reference
- **Model Codes:** P1-P20, IN1-IN20, CO1-CO20, DE1-DE20, RE1-RE20, SY1-SY20
- **Quality Standard:** 9.0/10 minimum for application
- **Validation:** Oct 29, 2025 Base120 specification
## Success Criteria
**Effective transformation application achieves:**
- ✅ Clear process followed
- ✅ Appropriate transformation selected
- ✅ Insights generated (not just analysis)
- ✅ Actionable outputs
- ✅ Documented reasoning
**Application fails if:**
- ❌ Wrong transformation chosen
- ❌ Process skipped/rushed
- ❌ No insights emerged
- ❌ Can't act on results
- ❌ Reasoning not documentedRelated Skills
test-and-fix-workflow
Automated workflow for running tests and fixing failures systematically. Use when implementing the mandatory test workflow or fixing code quality issues. Keywords - testing, debugging, workflow, failures, systematic fixes.
team-workflows
Team collaboration patterns - shared configs, standards, onboarding
tdd-workflows-tdd-refactor
Use when working with tdd workflows tdd refactor
tdd-workflows-tdd-red
Generate failing tests for the TDD red phase to define expected behavior and edge cases.
tdd-workflows-tdd-green
Implement the minimal code needed to make failing tests pass in the TDD green phase.
spec-workflow
Standard software engineering workflow for requirement analysis, technical design, and task planning. Use this skill when developing new features, complex architecture designs, multi-module integrations, or projects involving database/UI design.
spec-workflow-orchestrator
Orchestrate comprehensive planning phase from ideation to development-ready specifications using 4 specialized agents
python-workflow
Python project workflow guidelines. Triggers: .py, pyproject.toml, uv, pip, pytest, Python. Covers package management, virtual environments, code style, type safety, testing, configuration, CQRS patterns, and Python-specific development tasks.
python-workflow-development
Develop Python scripts and modules for building AI workflows and integrations. Use when coding data ingestion, transformation, analysis, and automation pipelines in pilot projects requiring Python automation.
Proactive Review Workflow
Automate multi-pass code review across quality, security, and performance dimensions
prisma-workflow
Prisma workflow for schema changes, migrations, and common pitfalls in this repo.
parallel-workflows
Optimizes parallel execution of multiple tasks. Use when user mentions 並列で実行, 同時にやって, まとめてやって, run in parallel, do these together. Do NOT load for: 単一タスク, 順次実行が必要な作業, 依存関係のあるタスク.