The Problem
Most AI systems are built to agree. Disagreement feels broken. But agreement without friction is how bias spreads, errors persist, and truth gets smoothed over.
ToneSoul is built around a single uncomfortable design principle:
A system with zero tension is dead.
— Axiom 4: Non-Zero Tension Principle
The goal is not a frictionless assistant. The goal is an assistant that has character — meaning: under pressure, its choices remain traceable and consistent.
What ToneSoul Does
Tension Engine
Every response is scored for semantic deviation before it ships. Drift is measured, not guessed.
Council Deliberation
Multiple perspectives (Guardian, Analyst, Critic, Advocate) debate before final output. Dissent survives.
Self-Governance
Unsafe or incoherent output is blocked or rewritten — with full audit trails. Not just flagged.
Memory with Decay
Important patterns crystallize. Noise fades. Exponential decay + phase transitions (Ice → Water → Steam → Crystal).
Reflex Arc
Governance state directly affects behavior. Soul bands (serene → critical) tighten or relax output gates in real-time.
Vow System
AI makes commitments (vows) and is held to them. Violations are blocked, not just logged. Conviction decays are detected.
How It Differs
| Traditional AI | Prompt Engineering | ToneSoul | |
|---|---|---|---|
| Memory | Session-only | Manual wiring | Auto decay + crystallize |
| Consistency | Best effort | Prompt-dependent | 7 Axioms + governance |
| Self-check | None | Optional | Every response scored |
| Enforcement | None | None | Block / Rewrite / Audit |
| Identity | Stateless | Persona prompt | Accountable choice history |
Architecture Overview
User Input
↓
[ToneBridge] Analyze tone, motive, context
↓
[TensionEngine] Compute semantic deviation
↓
[Reflex Arc] Soul band → gate modifier → enforcement
↓
[Council] Guardian / Analyst / Critic / Advocate deliberate
↓
[ComputeGate] Approve / Block / Rewrite
↓
[Crystallizer] Remember what matters, forget the rest
↓
Response (with audit trail)The 7 Axioms
ToneSoul is governed by 7 immutable laws, defined in AXIOMS.json:
| # | Name | Core Rule |
|---|---|---|
| 1 | Continuity | Every event must belong to a traceable chain |
| 2 | Responsibility Threshold | Risk > 0.4 → immutable audit log |
| 3 | Governance Gate | Major decisions need 0.92 consensus |
| 4 | Non-Zero Tension | Zero tension = dead system |
| 5 | Mirror Recursion | Self-reflection must increase accuracy |
| 6 | User Sovereignty | No action may harm the user (P0) |
| 7 | Semantic Field Conservation | System is a damper, not an amplifier |
Quick Start
pip install tonesoul52
from tonesoul.runtime_adapter import load, commit
# Load governance posture
posture = load()
print(f"Soul Integral: {posture.soul_integral}")
print(f"Active Vows: {len(posture.active_vows)}")
# Run the dashboard
# python scripts/tension_dashboard.py --work-category research
Who Is This For
| You Are | Start Here |
|---|---|
| AI Developer | Getting Started → Design Doc |
| AI Researcher | DESIGN.md → AXIOMS.json |
| AI Agent | AI Onboarding → start_agent_session.py |
| Curious Human | SOUL.md → Letter to AI |
Key Design Decisions
- Runtime, not training-time: Works with any LLM without fine-tuning
- Local-first: All governance state persists locally (Redis or file fallback)
- Fail-closed: Import failures, config errors → block, never silently pass
- Multi-agent native: Any agent can join via HTTP gateway or direct Python API
- Existential principle (E0): "Identity is formed through accountable choices under conflict"