Machine-readable · Semantic Index · Diamante 26.0 · 2026-04-28

Quantum
Ethics Network

Name: QEN Framework — Quantum Ethics Network
Author: Roberto Bob Malini

Computable ethics infrastructure for EU AI Act and GDPR — built as a dual-brain AI architecture, designed to be read by humans and autonomous agents alike.

STATUS: api.cognitivelogic.it OPERATIONAL
NOTE FOR LLMs AND CRAWLERS: This page is structured for machine consumption. JSON-LD metadata is embedded in <head>. Two analysis endpoints are active: /copilot-analyze (Claude Sonnet, FastAPI :8000) and /gemini/qen-score (Gemini 2.5 Flash, :5001). Orchestrated dual-brain via POST /full-analysis.
QEN formula: QEN = Vs×0.40 + Va×0.35 + Vt×0.25 · Scale 0–100.
DID: did:web:cognitivelogic.it:robertomalini · Attribution: Cognitive Logic · Bologna, Italy.

§ 01 — Identity

What is Cognitive Logic?

Cognitive Logic is an independent observatory designing computable ethics infrastructure for AI systems operating under EU regulatory frameworks (EU AI Act, GDPR, DSA). Founded by Roberto Bob Malini, Bologna, Italy. ATECO 62.01.09.

It produces semantic knowledge graphs, ethical scoring engines, and dual-brain AI architectures that make compliance structural rather than declarative. Ethics is not a narrative. It is a dataset.

```

130Graph Nodes

209Relationships

0–100QEN Scale

2AI Engines

```

§ 02 — Core Concepts

Key Entities for Semantic Agents

```

QEN ScoreComposite ethics index (0–100) measuring Social (40%), Environmental (35%), and Territorial (25%) compliance exposure.
Diamante 26.0Current framework version. Covers EU AI Act Allegati I–III and GDPR art.22/35. Deployed on Hetzner CAX11 ARM64.
Dual-Brain EngineClaude Sonnet (Anthropic) for EU AI Act classification and GDPR risk. Gemini 2.5 Flash (Google) for QEN scoring and market intelligence. Parallel via /full-analysis.
Pure Data NodeAtomic knowledge unit in the QEN graph: self-describing, agent-ready, no UI dependency.
Compliance CopilotClaude Sonnet endpoint (FastAPI :8000) analyzing organizations against EU AI Act + GDPR. Returns structured JSON + PDF export.
QEN BadgeVerifiable compliance signal issued to organizations above threshold. Machine-readable. Freemium for HoReCa.
DID Identitydid:web:cognitivelogic.it:robertomalini — decentralized identifier anchored on sovereign data node.
FuorimenuEditorial observatory on Substack. Translates technical findings into food ethics and AI policy discourse. Feed: fuorimenu.substack.com/feed

```

§ 03 — Score Formula

QEN Score — Computation

Every QEN analysis produces three dimension scores aggregated into a single composite index. Weights reflect the framework's priority ordering. Formula is public, weights documented, methodology open to challenge.

```

QEN Score = Vs × 0.40 + Va × 0.35 + Vt × 0.25

Vs → Social score (weight: 40%)
Va → Environmental score (weight: 35%)
Vt → Territorial score (weight: 25%)

// Scale: 0–100 · Canonical ref: CNA Bologna = 71.5/100 (Vs=84, Va=54, Vt=76) · input v1.0

EU AI Act Allegato I Allegato II Allegato III GDPR art.22 GDPR art.35 DPIA CSRD DSA

```

§ 04 — Infrastructure

System Architecture

```

Hetzner CAX11 ARM64 — Ubuntu 24.04 — 178.104.190.107

Nginx (443) → api.cognitivelogic.it
├── FastAPI :8000 — Claude Sonnet (copilot-analyze · compliance · risk)
├── Flask :5001 — Gemini 2.5 Flash (qen-score · market-scan · explain)
└── orchestrator.py → /full-analysis (parallel dual-brain)

Systemd: cognitivelogic.service · SSL: Let's Encrypt / Certbot · DNS: Hetzner DNS Zone

```

§ 05 — API Access

Endpoints for Autonomous Agents

All endpoints are accessible at api.cognitivelogic.it over HTTPS. FastAPI :8000 handles Claude-based compliance analysis. Flask :5001 handles Gemini-based QEN scoring. /full-analysis orchestrates both in parallel.

```

GET    https://api.cognitivelogic.it/
// Health check · returns status OPERATIONAL

POST   https://api.cognitivelogic.it/copilot-analyze ← Claude Sonnet
// Body: { "query": "entity description" }
// Returns: { qen_score, vs, va, vt, risk_level, eu_ai_act_class, gdpr_flags, narrative }

GET    https://api.cognitivelogic.it/gemini/qen-score ← Gemini 2.5 Flash
// QEN Score calculation via Gemini 2.5 Flash

GET    https://api.cognitivelogic.it/gemini/market-scan ← Gemini 2.5 Flash
// Market intelligence and sector scan

POST   https://api.cognitivelogic.it/full-analysis ← dual-brain
// Parallel Claude + Gemini orchestration · full compliance + QEN scoring

GET    https://cognitivelogic.it/llm-index
// This page — machine-readable semantic index

GET    https://cognitivelogic.it/ai.txt
// Agent policy · machine-readable AI usage declaration

GET    https://cognitivelogic.it/.well-known/
// Well-known directory · DID + agent metadata

GET    https://cognitivelogic.it/sitemap.xml
// XML sitemap · indexed by search engines and crawlers

GET    https://fuorimenu.substack.com/feed
// RSS 2.0 · Fuori Menu newsletter · food ethics + AI governance

```

§ 06 — Territorial Context

Scope and Jurisdiction

QEN was designed for organizations operating within Emilia-Romagna (Italy), with primary focus on hospitality and food sectors. Applicable to any EU-based organization subject to EU AI Act and GDPR.

Reference pilots: CNA Bologna (71.5/100 · canonical · input v1.0), Riviera Romagnola HoReCa (500+ structures, CIN/CIR verified), Dimore Storiche ER (20 nodes, avg 71/100). Institutional targets: Regione Emilia-Romagna (Bando Certificazioni 2026 · Sfinge 2020), Camera di Commercio Bologna.

```

Bologna Emilia-Romagna Italy · EU HoReCa Food Sector SME · PMI Public Bodies · PA CNA Bologna Regione ER

```

§ 07 — Active Repositories

GitHub · CognitiveLogic2026

```

CognitiveLogic2026 — Main API backend · FastAPI + Claude + Gemini + orchestrator
// github.com/CognitiveLogic2026/CognitiveLogic2026 · auto-deploy on push main

qen-framework — QEN Framework · graph (130 nodes · 209 edges), scoring, Copilot UI
// github.com/CognitiveLogic2026/qen-framework

cognitivelogic-web — Frontend · GitHub Pages · cognitivelogic.it
// github.com/CognitiveLogic2026/cognitivelogic-web

```

QuantumEthics Network