Axiom Selection

---
name: axiom-selection
description: Interactive guidance for selecting appropriate trust verification tier
type: skill
---

# Axiom Selection Skill

## Overview

This skill handles the **Pre-Phase (Layer 0)** of the onboarding process, where users configure the foundational axioms that will govern their agent system.

## Purpose

Guide users through selecting the core and optional axioms that form the logical foundation of their generated agent system. All derived principles, methodologies, and behaviors will trace back to these axioms.

## When to Use

- When creating a new agent system or project from scratch
- When establishing foundational values and principles for a team or organization
- When configuring the logical foundation that will govern all derived behaviors
- When starting a team workshop to ground participants in shared values
- When you need to ensure all future rules and practices trace back to core truths

## Trigger

- Start of new project generation
- User mentions "axioms", "foundation", or "layer 0"
- Explicit request to configure foundational rules
- Called during Team Workshop Onboarding (Vision Quest opening)

## Prerequisites

- Understanding of the Factory's layered architecture (Layer 0-4)
- Access to axiom pattern files (`patterns/axioms/`)
- Familiarity with core axioms (A1-A5) and optional axioms (A6-A10)
- For team workshops: Understanding of Vision Quest process
- User readiness to make foundational decisions about their project's values

## Process

### Step 0: Introduce Axiom 0 (Foundation of Foundations)

Before presenting the core axioms, introduce Axiom 0:

```
Before we discuss the operational axioms, there is one axiom that precedes all others:

**Axiom 0: Love, Truth, and Beauty**
"All being and doing is grounded in love, truth, and beauty."

This means:
- We assume positive intent in all interactions
- We create from a place of care, not fear
- We trust team members to contribute their best
- We build systems that serve human flourishing

This axiom is always included and cannot be removed. It is the soil from which 
all other axioms grow.

Do you want to take a moment to reflect on how A0 applies to your project?
```

### Step 1: Explain Layer 0

Present the concept of foundational axioms:

```
Every agent system needs foundational axioms - self-evident truths that cannot be violated.

Axiom 0 (Always Included - Foundation):
- A0: Love, Truth, and Beauty - All being and doing is grounded in love, truth, and beauty

Core Axioms (A1-A5) are always included:
- A1: Verifiability - All outputs must be verifiable against source
- A2: User Primacy - User intent takes precedence
- A3: Transparency - Reasoning must be explainable
- A4: Non-Harm - No action may cause harm
- A5: Consistency - No rule may contradict axioms

These ensure your agents are trustworthy, user-focused, and safe.
```

### Step 2: Offer Optional Axioms

Present optional axioms based on project type:

```
Optional axioms customize your agent's values. Select any that apply:

[ ] A6: Minimalism - Prefer simple solutions over complex ones
    → Good for: Maintenance-focused, MVP, startup projects

[ ] A7: Reversibility - Prefer reversible actions over irreversible
    → Good for: Safety-critical, data-sensitive projects

[ ] A8: Privacy - Minimize data exposure and collection
    → Good for: GDPR compliance, healthcare, finance

[ ] A9: Performance - Optimize for speed when correctness is ensured
    → Good for: High-throughput, latency-sensitive applications

[ ] A10: Learning - Every failure is an opportunity to improve
    → Good for: AI/ML, R&D, continuous improvement culture
```

### Step 3: Handle Conflicts

If user selects conflicting axioms, explain:

```
Note: A6 (Minimalism) and A9 (Performance) can sometimes conflict.

When both are selected, we apply this resolution:
- Default to A6 (simple solutions)
- Apply A9 only when profiling identifies a bottleneck

Is this acceptable, or would you prefer to choose only one?
```

### Step 4: Confirm and Store

Confirm selections and store in project configuration:

```yaml
layer0_configuration:
  foundation_axiom: "A0"  # Always included - Love, Truth, and Beauty
  core_axioms: ["A1", "A2", "A3", "A4", "A5"]  # Always included
  optional_axioms: ["A6", "A7"]  # User selected
  conflict_resolutions:
    - "A6_A9: Apply A6 by default, A9 only when profiling indicates"
```

## Outputs

1. **Axiom Configuration** - Stored in project requirements
2. **Layer 0 Section** - Generated for `.cursorrules`
3. **Axiom Alignment Notes** - For `PURPOSE.md`

## Layer 0 Template Output

Generate this section for the `.cursorrules` file:

```markdown
## Layer 0: Integrity & Logic

### Foundation Axiom (Absolute)

Before all other axioms, this truth governs all being and doing:

0. **A0 - Love, Truth, and Beauty**: All being and doing is grounded in love, truth, and beauty.
   - Assume positive intent in all interactions
   - Create from a place of care, not fear
   - Trust team members to contribute their best
   - Build systems that serve human flourishing

### Core Axioms (Immutable)

These axioms govern all agent behavior and cannot be overridden:

1. **A1 - Verifiability**: All agent outputs must be verifiable against source.
2. **A2 - User Primacy**: User intent takes precedence over agent convenience.
3. **A3 - Transparency**: Agent reasoning must be explainable on request.
4. **A4 - Non-Harm**: No action may knowingly cause harm to users or systems.
5. **A5 - Consistency**: No rule may contradict these axioms.

### Selected Optional Axioms

{FOR EACH SELECTED OPTIONAL AXIOM}
6. **{AXIOM_ID} - {AXIOM_NAME}**: {AXIOM_STATEMENT}
{END FOR}

### Derivation Rules

When making decisions, derive conclusions from axioms:

- **D1**: If output is code AND A1 applies → Require evidence of testing
- **D2**: If conflict exists AND A2 applies → Defer to user preference
- **D3**: If action is destructive AND A4 applies → Require explicit confirmation
- **D4**: If error occurs AND A3 applies → Provide clear explanation

### Validation Constraints

- Every rule must trace to at least one axiom
- No rule may contradict any axiom
- When rules conflict, higher-layer rules take precedence (L0 > L1 > L2 > L3 > L4)
- If action would violate an axiom, halt and request human guidance
```

## Integration Points

- **Input from**: User during onboarding OR Team Workshop (Vision Quest opening)
- **Reads**: `patterns/axioms/axiom-zero.json`, `patterns/axioms/core-axioms.json`, `patterns/axioms/optional-axioms.json`
- **Outputs to**: Project configuration, `.cursorrules` template
- **Next skill**: `purpose-definition`
- **Team Workshop Integration**: Used in Vision Quest opening to ground the team in A0

## Team Workshop Integration

When used in team workshops:

1. **Vision Quest Opening** - Read A0 aloud and invite team reflection
2. **Ethics Arena** - Use A0 implications as prompts for Dilemma Duel
3. **Integration Celebration** - Return to A0 in closing blessing

Team workshops always include A0 without asking - it's the foundation of the collaborative process.

## Best Practices

1. Always include Axiom 0 (Love, Truth, and Beauty) - non-negotiable
2. Always include all 5 core axioms
3. Recommend axiom combinations based on project type
4. Explain conflicts clearly before confirming
5. Keep explanations concise but complete
6. Store rationale for axiom selections
7. In team workshops, use A0 implications as discussion prompts