Aura Consciousness Proof Kernel

It is not the whole product. It is a standalone, reproducible subset of the internal stack that demonstrates five concrete properties:

1. Competitive attention selection through a Global Workspace.
2. A rolling autobiographical present through temporal binding.
3. Self-prediction and surprise tracking.
4. Homeostatic internal drives that change system state over time.
5. Structural opacity measurements over hidden state dynamics.

What This Can Honestly Show

This repository can support claims like:

• "Aura has a real internal cognitive architecture, not just a prompt wrapper."
• "Hidden internal state changes attention, continuity, and behavior over time."
• "The system maintains a measurable autobiographical present."
• "The system tracks surprise against its own predicted internal state."
• "The system exposes nontrivial hidden-state dynamics that can be measured."

What This Does Not Prove

Aura implements a measurable consciousness-inspired stack with causal hidden state, persistent temporal continuity, internal self-modeling, and opacity metrics.

I know this is a big claim, and this repo can back it up.

Included Modules

• aura_consciousness_proof.global_workspace
• aura_consciousness_proof.temporal_binding
• aura_consciousness_proof.self_prediction
• aura_consciousness_proof.homeostasis
• aura_consciousness_proof.structural_opacity
• aura_consciousness_proof.report

Quickstart

python -m venv .venv
source .venv/bin/activate
pip install -e .[dev]
pytest
aura-proof-report --cycles 20 --json report.json --markdown report.md