Metadata-Version: 2.4
Name: aria-engagEd
Version: 1.0.0
Summary: ARIA behavioural‑intelligence substrate (OEM edition)
Author: EngagEd / ARIA Behavioural Intelligence
License: ARIA COMMERCIAL OEM LICENSE
        
        Version 1.0 | August 2026 | ARIA EngagEd Ltd
        
        ===================================
        
        1. GRANT OF LICENSE
        
        ARIA EngagEd Ltd ("Licensor") grants Licensee a non-exclusive, 
        non-transferable, perpetual license to use the ARIA substrate 
        ("Software") solely for integration into Licensee's products 
        and services.
        
        2. RESTRICTIONS
        
        Licensee may NOT:
        - Sublicense or redistribute the Software
        - Remove or obscure proprietary notices
        - Reverse engineer or decompile
        - Use for competing behavioral intelligence products
        - Audit trail tampering or cryptographic verification bypass
        
        3. AUDIT & COMPLIANCE
        
        Licensee shall permit annual audits of cryptographic audit trails
        and regulatory compliance via Ed25519 signature verification.
        
        4. LIABILITY LIMITATION
        
        ARIA is supplied "as is." Licensor disclaims all warranties.
        In no event shall Licensor be liable for indirect, incidental,
        or consequential damages arising from use.
        
        5. TERM & TERMINATION
        
        License term: Annual, renewable. Licensor may terminate for
        material breach with 30 days' notice.
        
        6. GOVERNING LAW
        
        This agreement is governed by the laws of England and Wales.
        
        ===================================
        
        For licensing inquiries: sales@ariaengaged.co.uk
        
Project-URL: Homepage, https://github.com/ARIA-EngagEd-Ltd/ARIA
Project-URL: OEM Documentation, https://github.com/ARIA-EngagEd-Ltd/ARIA/blob/main/docs/OEM/README-OEM.md
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: License :: Other/Proprietary License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: PyJWT>=2.0
Requires-Dist: requests>=2.28
Requires-Dist: cryptography>=50.0
Requires-Dist: pyyaml>=6.0
Requires-Dist: fastapi>=0.104.0
Requires-Dist: uvicorn[standard]>=0.24.0

# ARIA — Behavioural‑Intelligence Substrate
Version: 1.0.0

ARIA is a deterministic, 7-layer behavioral-intelligence substrate designed for OEM
integration. It ingests events, normalizes them, verifies identity, assembles operational
context, evaluates policies, synthesizes remediation patches, and gates execution with
human approval. Processes metadata‑only inputs and returns stable behavioral decisions
with full audit traceability.

This repository contains the complete 7-layer ARIA engine, policy evaluator, patch
synthesizer, execution gate, audit system, OEM documentation pack, and packaging
configuration.

---

## 7-Layer Architecture

**Layer 1: Ingestion & Normalization**
- Raw event ingestion from diverse sources
- Normalization to observation primitives

**Layer 2: Convergence**
- Time-aligned canonical event construction
- Schema validation and format standardization

**Layer 3: Identity Verification**
- Cryptographic identity verification (JWT, SPIFFE, OIDC)
- Key management and rotation

**Layer 4: Operational Context**
- Entity resolution and relationship mapping
- Baseline computation and anomaly detection
- Operational graph assembly

**Layer 5: Policy Evaluation**
- Behavioral policy evaluation engine
- Decision generation with confidence scores
- Extensible policy framework

**Layer 6: Patch Synthesis**
- Automated remediation patch generation
- JSON patch construction for policy violations

**Layer 7: Execution Gating**
- Human approval requirement for execution
- Signature verification
- Audit trail completion

## Features

- Deterministic 7-layer behavioral-decision pipeline
- Metadata‑only processing (no raw content ingestion)
- Cryptographic identity verification
- Suppression logic and anomaly detection
- Full audit trail with decision provenance
- Extensible policy framework for domain verticals
- OEM documentation pack
- Python package (pyproject.toml)
- Minimal external dependencies

---

## Repository Structure

```
ARIA/
  ├── VERSION
  ├── CHANGELOG.md
  ├── pyproject.toml
  ├── docs/OEM/
  │   ├── README-OEM.md
  │   ├── INTEGRATION-GUIDE.md
  │   ├── API-SURFACE.md
  │   ├── AUDIT.md
  │   └── LICENSE-OEM.txt
  └── src/aria/
      ├── layer1/        # Ingestion & Normalization
      ├── layer2/        # Convergence
      ├── layer3/        # Identity Verification
      ├── layer4/        # Operational Context
      ├── layer5/        # Policy Evaluation
      ├── layer6/        # Patch Synthesis
      ├── layer7/        # Execution Gating
      ├── regulatory/    # Audit & Reporting
      ├── api/aria_api.py
      └── verticals/     # Domain-specific extensions (Financial, etc.)
```

---

## Dependencies

ARIA requires minimal external dependencies:

- `PyJWT>=2.0` — JWT token verification for Layer 3 identity verification
- `requests>=2.28` — HTTP client for OIDC provider communication

---

## Installation

ARIA requires Python >=3.10.

Install from source:

```
pip install .
```

---

## Usage

```python
from aria.api import AriaAPI

engine = AriaAPI()
signals = engine.run(metadata)
```

---

## OEM Documentation

See `docs/OEM/` for:

- Integration Guide
- API Surface
- Audit System
- Versioning
- Changelog

---

## Versioning

ARIA follows semantic versioning.
OEMs must pin ARIA to:

```
>=1.0.0 <2.0.0
```

---

## License

See `docs/OEM/LICENSE-OEM.txt`.
