Apply structured thinking frameworks to decisions, problems, and planning.

When This Triggers

"Help me think through X"
"What mental model should I use?"
"Apply first principles to this"
"Do a pre-mortem on this plan"
"I need to analyze this decision"
Any specific model name (SWOT, 5 Whys, etc.)

Workflow

Step 1: Run Mental Models Scanner

Run the script to get all available models:

python 00-system/mental-models/scripts/select_mental_models.py --format brief

This returns JSON with all 59 models across 12 categories.

Optional filters:

# Filter by category
python 00-system/mental-models/scripts/select_mental_models.py --category cognitive --format brief

# List format (names only, grouped by category)
python 00-system/mental-models/scripts/select_mental_models.py --format list

Step 2: Identify Context and Offer Options

Based on user's situation, offer 2-3 relevant models:

For decisions: Decision Matrix, Pre-Mortem, Cost-Benefit, Inversion For problems: First Principles, Root Cause (5 Whys), Fishbone For planning: Scenario Planning, Stakeholder Mapping, OKR For creativity: Design Thinking, SCAMPER, Lateral Thinking For risk: Pre-Mortem, Force Field, Red Team, Black Swan For communication: Pyramid Principle, BLUF, Steel Manning For learning: Feynman Technique, Deliberate Practice

Present options with brief descriptions (3-7 words each).

Step 3: Load Specific Model File

After user selects, load the individual model file:

File structure: 00-system/mental-models/models/{category}/{model-slug}.md

| Cognitive | models/cognitive/first-principles.md, inversion.md, etc.

| Collaborative | models/collaborative/six-thinking-hats.md, mece.md, etc.

| Diagnostic | models/diagnostic/pre-mortem.md, five-whys.md, etc.

| Strategic | models/strategic/scenario-planning.md, ooda-loop.md, etc.

| Analytical | models/analytical/decision-matrix.md, swot-analysis.md, etc.

| Creative | models/creative/design-thinking.md, scamper.md, etc.

| Operational | models/operational/kanban-thinking.md, okr-framework.md, etc.

| Validation | models/validation/hypothesis-testing.md, red-team-analysis.md, etc.

| Time & Resource | models/time-resource/eisenhower-matrix.md, opportunity-cost.md, etc.

| Communication | models/communication/pyramid-principle.md, bluf.md, etc.

| Learning | models/learning/feynman-technique.md, deliberate-practice.md, etc.

| Probability & Risk | models/probability-risk/expected-value.md, black-swan-awareness.md, etc.

Example:

User picks: "First Principles + Pre-Mortem"

AI loads:
→ Read: 00-system/mental-models/models/cognitive/first-principles.md
→ Read: 00-system/mental-models/models/diagnostic/pre-mortem.md

Step 4: Apply Model Questions

Use the question templates from the loaded model file to guide the user through structured thinking.

Each model file contains:

Purpose: What the model does
When to Use: Best situations for this model
Questions to Ask: Ready-to-use prompts
Process: Step-by-step application
Output: What you get from using it

Keep it collaborative - this is a conversation, not an interrogation.

Quick Reference

12 Categories (59 Models):

Cognitive - First Principles, Systems Thinking, Lateral Thinking, Inversion, Second-Order Thinking, Analogous Reasoning
Collaborative - Six Hats, MECE, Stakeholder Mapping, Devil's Advocate
Diagnostic - 5 Whys, Fishbone, Pre-Mortem, Force Field, Fault Tree
Strategic - Scenario Planning, OODA, Jobs to Be Done, Blue Ocean, PESTLE, Porter's Five Forces
Analytical - Decision Matrix, SWOT, Cost-Benefit, Pareto, Assumption Testing, Sensitivity Analysis
Creative - Design Thinking, SCAMPER, Morphological, Random Entry, Constraint Removal
Operational - Kanban, Value Stream, OKR, Lean Canvas, Theory of Constraints
Validation - Hypothesis Testing, Prototyping, Red Team, A/B Testing
Time & Resource - Eisenhower Matrix, Time Boxing, Opportunity Cost, Sunk Cost, Resource Mapping
Communication - Pyramid Principle, BLUF, Situation-Complication-Resolution, Steel Manning
Learning - Feynman Technique, Spaced Repetition, Deliberate Practice, T-Shaped Skills
Probability & Risk - Expected Value, Margin of Safety, Black Swan, Bayesian Updating, Regret Minimization

Notes

Always offer choice, never prescribe
Run script first to see all available models
Load individual model files only after user selects
Combine models when appropriate (e.g., First Principles + Pre-Mortem)
Adapt formality to user's context

安装