Aurora Workflow Orchestration (AWO)

Code under Apache 2.0 · Documentation under CC BY 4.0

Aurora Workflow Orchestration (AWO)
A formal method for reproducible AI-assisted research
• Falsifiability • Provenance • Attestation • Auditability
• Works manually or via CRI-CORE automation
• Every artifact is signed, structured, and verifiable

AWO is a reproducibility framework for AI-assisted research — turning every run into a verifiable scientific artifact.

---

Overview

Aurora Workflow Orchestration (AWO) defines how AI-assisted research can be made reproducible and auditable.
It translates reasoning steps, decisions, and evidence into structured files that anyone can verify.

AWO is the methodology layer — it governs how reproducibility works.
CRI-CORE is the execution layer — it automates that governance.
Together they form a single, evidence-based research system.

Idea → Manifest → Run → Audit → Attestation → Archive  
    |________ Governance Rules (AWO) ________|
                         ↓
             |_____ Enforcement (CRI-CORE) ____|

---

Repository Structure

This repository contains the complete AWO governance layer — the foundation for reproducible, falsifiable AI research.

Core Documents

• Whitepaper (PDF) — Rationale and design philosophy
• Method Spec (PDF) — Normative rules for compliance
• Adoption Guide (PDF) — Step-by-step onboarding

Governance & Evidence

• GOVERNANCE_SUMMARY.md — Compliance and attestation records
• ROLE_ATTESTATION.md — Role-level verification summary
• AWO_Compliance_Report.md — Signed compliance certification

Design & Provenance

• Architecture Decision Records (/decisions/ADR-0001 → ADR-0018)
• Validation Schemas (/schemas/) and Templates (/templates/)
• Reproducible Runs (/runs/) with frozen manifests and signed approvals

All content is cryptographically verified through SHA256SUMS.txt
and governed by ADR-0015 → ADR-0018 under the Aurora Research Initiative.

---

Why AWO Exists

AI now generates discoveries faster than science can verify them.
The result is insight without integrity.

AWO is the countermeasure — a governance layer that forces every idea to prove itself before it earns the name “knowledge.”

Core Principles

1. Falsifiability First – every claim must define how it can fail.
2. Human-in-the-Loop Rigor – AI output remains a hypothesis until verified.
3. Immutable Provenance – every artifact is signed, hashed, and auditable.
4. Transparent Governance – reproducibility replaces reputation.

AWO turns the scientific method into a living protocol for evidence.

---

Quick Start

You don’t need special tools or a PhD in YAML to use AWO.
If you can commit to a Git repo, you can commit to reproducibility.

1. Clone or fork this repository

   git clone https://github.com/Waveframe-Labs/Aurora-Workflow-Orchestration.git
   cd Aurora-Workflow-Orchestration

2. Open the Adoption Guide
   → docs/AWO_Adoption_Guide.md
   It walks you through creating your first falsifiability manifest and recording an attested run.

3. Inspect a verified example
   Browse /runs/ to see how manifests, approvals, and logs form a complete provenance chain.

---

Adoption Tiers

Tier	Audience	What It Includes
Minimum	Individuals	Manual logs + falsifiability manifests
Standard	Small teams	CI pipelines + attestation workflows
Full	Institutions	Automated reproducibility via CRI-CORE

---

Example Projects Using AWO

Project	Domain	Mode of Use
Waveframe v4.0	Cosmology	Manual orchestration with falsifiability logs and ADRs
Societal Simulator	Systems modeling	Demonstrates reproducibility without automation
CRI-CORE	Research runtime	Automated orchestration and provenance enforcement

---

Common Questions

Is AWO useful if I work alone?

Absolutely. AWO scales from solo projects to full institutions. If you’re a lab of one with coffee and conviction, it still works.

Do I need CRI-CORE to use AWO?

No. AWO is fully functional on its own. CRI-CORE just automates what you can already do manually.

Does AWO replace peer review?

Not at all. It strengthens it — by ensuring every claim and artifact is traceable before publication.

Can I publish AWO-based research?

Yes. Include the AWO concept DOI (10.5281/zenodo.17013612) in your Methods or reproducibility statement.

---

🙅‍♂️ What AWO Is Not

• ❌ A software package — it’s a method with optional automation.
• ❌ A belief system — it’s governance, not gospel.
• ❌ Dependent on institutions — reproducibility is the credential.
• ❌ Too heavy for individuals — AWO scales down cleanly to one researcher.

---

Version Boundary

AWO v1.2.1 (Documentation and Accessibility Update) marks the finalization of the AWO methodology under Waveframe Labs governance.
Future updates appear only as errata or citation additions.

Canonical DOI: 10.5281/zenodo.17013612

---

Citation

If you reference or build upon AWO, please cite using the concept DOI.

APA

Wright, S. C. (2025). Aurora Workflow Orchestration (AWO): A formal framework for reproducible AI-assisted research.
Waveframe Labs / Aurora Research Initiative. https://doi.org/10.5281/zenodo.17013612

BibTeX

@software{wright_aurora_workflow_orchestration_2025,
  author       = {Wright, Shawn C.},
  title        = {Aurora Workflow Orchestration (AWO): A formal framework for reproducible AI-assisted research},
  year         = {2025},
  version      = {1.2.1},
  institution  = {Waveframe Labs / Aurora Research Initiative},
  license      = {Apache-2.0 (code), CC BY 4.0 (docs)},
  url          = {https://github.com/Waveframe-Labs/Aurora-Workflow-Orchestration},
  doi          = {10.5281/zenodo.17013612}
}

---

Licensing

• Code → Apache 2.0
• Documentation → CC BY 4.0

---

Repository Status

This repository is an archival reference artifact — stable, verifiable, and citable.
All new runtime development continues in CRI-CORE, which automates AWO’s governance logic.

---

Integrity and Verification

This repository maintains a cryptographic integrity registry (SHA256SUMS.txt) at the root.
It is automatically rebuilt and committed by the Build root SHA256SUMS workflow.

Scope

• All core documents under /docs/ (whitepaper, spec, guide)
• All ADRs under /decisions/
• Compliance and governance files at root
• Current attested runs under /runs/

To verify locally:

sha256sum --check SHA256SUMS.txt

---

_{© 2025 Waveframe Labs · Independent Open-Science Research Entity}