README.md

ARI Cognitive Layer 0: LOGOS / ETHOS / PATHOS

Version: 1.0.0
Status: Design Phase Complete (Awaiting Council Vote)
Date: 2026-02-01

The mind behind the machine - Foundational consciousness layer providing algorithmic reasoning, emotional intelligence, and continuous learning to all of ARI.

---

What is Layer 0?

Layer 0 (Cognitive Foundation) is a new architectural layer beneath the Kernel that provides universal cognitive frameworks:

• LOGOS (Reason): Bayesian reasoning, expected value, Kelly Criterion, decision trees, systems thinking
• ETHOS (Character): Bias detection, emotional regulation, discipline, trading psychology
• PATHOS (Growth): CBT/DBT/ACT, Stoicism, deliberate practice, meta-learning, wisdom traditions

Result: Every Council member becomes cognitively enhanced - making decisions backed by probability theory, aware of biases, and continuously learning.

---

Quick Navigation

Architectural Decisions (ADRs)

1. ADR-009: Cognitive Layer Architecture - Why Layer 0, positioning, integration
2. ADR-010: Three-Pillar Framework - LOGOS/ETHOS/PATHOS structure
3. ADR-011: Knowledge Source Trust - 4-tier trust model, 5-stage validation
4. ADR-012: Council Specializations - Mapping members to frameworks
5. ADR-013: Cognitive API Design - API patterns, response formats
6. ADR-014: Learning Loop - 5-stage continuous improvement

Pillar Designs

1. LOGOS Pillar (35+ pages) - Bayesian, EV, Kelly, systems thinking, Taleb
2. ETHOS Pillar (30+ pages) - Trading psych, biases, emotion, discipline
3. PATHOS Pillar (35+ pages) - CBT/DBT/ACT, Stoicism, meta-learning, wisdom

Supporting Documentation

1. Council Cognitive Profiles - All 15 members' specializations
2. Implementation Roadmap - 7 phases, 16 weeks
3. Testing Strategy - 265+ tests, coverage targets
4. Knowledge Sources Catalog - 87 curated sources

---

Key Concepts

Three Pillars

LOGOS (λόγος) - Reason:

• Algorithmic decision-making under uncertainty
• Frameworks: Bayesian, Expected Value, Kelly Criterion
• Primary users: SCOUT, MINT, ATLAS, TRUE

ETHOS (ἦθος) - Character:

• Emotional intelligence and bias mitigation
• Frameworks: Trading psychology, bias detection, discipline
• Primary users: AEGIS, MINT, EMBER, SCOUT

PATHOS (πάθος) - Growth:

• Self-reflection, learning, and wisdom
• Frameworks: CBT/DBT/ACT, Stoicism, deliberate practice
• Primary users: BLOOM, PULSE, VERA, PRISM

Integration with Existing Architecture

Layer 6: INTERFACES     (CLI, Dashboard)
Layer 5: EXECUTION      (Daemon)
Layer 4: STRATEGIC      (Council, Arbiter, Overseer)
Layer 3: CORE AGENTS    (Core, Guardian, Planner, Executor)
Layer 2: SYSTEM         (Router, Storage)
Layer 1: KERNEL         (Gateway, Sanitizer, Audit, EventBus)
Layer 0: COGNITIVE      ← NEW (LOGOS + ETHOS + PATHOS)
         ↑
    Knowledge Streams (87 curated sources)

Key: Higher layers can import from Layer 0, Layer 0 is self-contained.

---

Quick Start (After Implementation)

Using Cognitive APIs

// Import specific functions
import { calculateExpectedValue } from './cognition/logos/expected-value.js';
import { detectCognitiveBias } from './cognition/ethos/bias-detector.js';
import { reflectOnOutcome } from './cognition/pathos/reflection.js';

// Calculate EV
const ev = await calculateExpectedValue({
  description: 'Should I invest?',
  outcomes: [
    { description: 'Success', probability: 0.60, value: 1000, confidence: 0.80 },
    { description: 'Failure', probability: 0.40, value: -500, confidence: 0.80 },
  ],
});
// Returns: { expectedValue: 400, recommendation: 'proceed', ... }

// Check for biases
const biases = await detectCognitiveBias(
  'I should invest because I read one positive article',
  { agent: 'MINT' }
);
// Returns: [{ bias: 'CONFIRMATION_BIAS', severity: 0.70, ... }]

// Reflect on outcome
const reflection = await reflectOnOutcome({
  action: 'Invested $1000',
  result: 'success',
  expectedValue: 400,
  actualValue: 600,
  timestamp: new Date().toISOString(),
});
// Returns: { insights: [...], principles: [...], nextActions: [...] }

---

Status

Design Phase: ✅ Complete (6 ADRs + 3 pillar designs + 6 supporting docs = ~350 pages)
Implementation Phase: ⏳ Pending (awaits Council vote)
Estimated Timeline: 16 weeks (4 months) for full deployment
Minimum Viable: 4 weeks (LOGOS prototype only)

---

Next Steps

1. Council Vote: Requires SUPERMAJORITY (9 of 15 members)
2. If Approved: Begin Phase 0 (Foundation, 2 weeks)
3. Prototype: Phase 1 (LOGOS, 2 weeks)
4. Evaluate: Does LOGOS improve decisions? Measure and decide whether to continue.
5. Full Build: Phases 2-7 (12 weeks)

---

Questions?

• Architecture: See ADR-009
• Security: See ADR-011
• Implementation: See Roadmap
• Testing: See Testing Strategy

---

Last Updated: 2026-02-01
Documentation Complete: ✅ All 21 deliverables finished
Ready for: Council vote → Implementation