assumption-validator

What this skill does

# Assumption Validator

> "Make hidden assumptions visible; make visible assumptions testable."

Assumptions are the invisible architecture of every decision. This skill provides a systematic methodology for surfacing, classifying, validating, and stress-testing the assumptions that underlie decisions, strategies, plans, and architectures. The output is a RISK-ASSESSMENT artifact that transforms assumption risk into actionable guidance.

---

## 1. Purpose

### Core Value Proposition

Every decision rests on assumptions. Some are explicit and acknowledged. Many are implicit, hidden in the reasoning chain. The most dangerous are structural—embedded in how we frame the problem itself. This skill makes the invisible visible and the untested testable.

### Capabilities

| # | Capability | Phase | Value |
|---|-----------|-------|-------|
| 1 | Extract explicit assumptions from subject documents | 2 | Capture what's acknowledged |
| 2 | Surface implicit assumptions through structured probing | 2 | Reveal hidden dependencies |
| 3 | Identify structural assumptions in problem framing | 2 | Expose framing blind spots |
| 4 | Classify assumptions by type and risk profile | 3 | Enable prioritization |
| 5 | Perform load-bearing analysis for criticality scoring | 3 | Focus on what matters |
| 6 | Map assumption dependencies | 3 | Understand cascade effects |
| 7 | Validate assumptions against available evidence | 4 | Assess confidence levels |
| 8 | Execute counterfactual analysis (what-if scenarios) | 4 | Stress-test under alternatives |
| 9 | Calibrate confidence with epistemic labels | 4 | Communicate uncertainty |
| 10 | Derive risks from unvalidated/invalid assumptions | 5 | Convert to actionable risks |
| 11 | Generate RISK-ASSESSMENT artifact | 5 | Standardized output |
| 12 | Provide go/no-go recommendation | 5 | Decision support |

---

## 2. When to Use

### Ideal Use Cases

| Scenario | Why Assumption Validation Matters |
|----------|-----------------------------------|
| **Pre-commitment decision review** | Before committing resources, surface what you're betting on |
| **Strategy validation** | Strategies often embed untested market/competitive assumptions |
| **Investment due diligence** | Financial decisions rest on projections built on assumptions |
| **Architecture Decision Records (ADRs)** | Technical choices assume certain constraints and capabilities |
| **Product direction pivots** | Pivots invalidate old assumptions; what new ones are introduced? |
| **Risk assessments** | Risks often emerge from assumption failures |
| **Merger & acquisition analysis** | Synergy assumptions are notoriously optimistic |
| **Go-to-market plans** | Market assumptions may not survive contact with reality |
| **Capacity planning** | Growth assumptions drive infrastructure decisions |
| **Vendor selection** | Vendor capabilities are often assumed, not verified |

### Anti-Patterns (When NOT to Use)

| Anti-Pattern | Why It's Ineffective | Better Alternative |
|--------------|----------------------|-------------------|
| **Analysis paralysis** | Validating every assumption delays action forever | Set validation_intensity: light |
| **Reversible decisions** | Over-analyzing easily reversible choices wastes effort | Just decide and iterate |
| **Well-established facts** | Questioning gravity wastes time | Focus on genuine uncertainties |
| **Time-critical emergencies** | Fires need extinguishing, not philosophy | Use judgment, then debrief |
| **Exploratory research** | Early exploration should generate assumptions, not validate them | Use research-interviewer first |

---

## 3. Parameters

| Parameter | Type | Required | Default | Description |
|-----------|------|----------|---------|-------------|
| `subject_type` | enum | **yes** | — | decision \| strategy \| plan \| architecture \| investment \| product_direction |
| `validation_intensity` | enum | no | standard | light \| standard \| rigorous |
| `assumption_depth` | enum | no | include_implicit | explicit_only \| include_implicit \| full_structural |
| `counterfactual_analysis` | boolean | no | true | Run what-if scenarios for top assumptions |
| `confidence_threshold` | number | no | 0.7 | Minimum confidence for "validated" status (0.0-1.0) |
| `time_horizon` | enum | no | all | short_term \| medium_term \| long_term \| all |

### Parameter Effects Matrix

| Parameter | Effect on Phase 2 | Effect on Phase 3 | Effect on Phase 4 |
|-----------|------------------|------------------|------------------|
| `validation_intensity: light` | 2 surfacing techniques | Top 3 load-bearing only | Evidence check only |
| `validation_intensity: standard` | 4 surfacing techniques | Top 5 load-bearing | Evidence + counterfactual |
| `validation_intensity: rigorous` | All surfacing techniques | All assumptions scored | Full validation battery |
| `assumption_depth: explicit_only` | Document scan only | — | — |
| `assumption_depth: include_implicit` | + Inversion, Five Whys | — | — |
| `assumption_depth: full_structural` | + Frame questioning | — | — |
| `time_horizon: short_term` | — | Filter to 0-6 months | — |
| `time_horizon: long_term` | — | Filter to 2+ years | — |

---

## 4. Five-Phase Workflow

### Workflow Overview

```
┌─────────────────────────────────────────────────────────────────────────────┐
│                         ASSUMPTION VALIDATOR                                 │
├─────────────────────────────────────────────────────────────────────────────┤
│                                                                              │
│  ┌──────────────┐    ┌──────────────┐    ┌──────────────┐                   │
│  │   PHASE 1    │    │   PHASE 2    │    │   PHASE 3    │                   │
│  │   Subject    │───▶│  Assumption  │───▶│Classification│                   │
│  │   Intake     │    │  Surfacing   │    │& Prioritize  │                   │
│  └──────────────┘    └──────────────┘    └──────────────┘                   │
│         │                   │                   │                            │
│         ▼                   ▼                   ▼                            │
│    [Framed         [Raw Assumption      [Prioritized                        │
│     Subject]        Inventory]           List]                              │
│                                                 │                            │
│                                                 ▼                            │
│                          ┌──────────────┐    ┌──────────────┐               │
│                          │   PHASE 5    │◀───│   PHASE 4    │               │
│                          │  Synthesis   │    │  Validation  │               │
│                          │& Risk Assess │    │& Stress Test │               │
│                          └──────────────┘    └──────────────┘               │
│                                 │                   │                        │
│                                 ▼                   ▼                        │
│                          [RISK-ASSESSMENT   [Validated/                     │
│                           CONTRACT-08]       Invalidated]                   │
│                                                                              │
└─────────────────────────────────────────────────────────────────────────────┘
```

---

### Phase 1: Subject Intake & Framing

**Purpose:** Understand what's being validated and establish clear boundaries.

**Steps:**

1. **Receive subject artifact**
   - Accept decision document, strategy brief, plan, ADR, or proposal
   - If verbal, request written summary or create one together
   - Note: Subject should be specific enough to identify assumptions

2. **Extract context metadata**
   - Subject type (decision, strategy, plan, architecture, investment, product_direction)
   - Time horizon (when does this need to succeed?)
   - Stakeholders (whose assumptions might be embedded?)
   - Decision reversibility (high-stakes or easily