DEMI Pipeline—Working Archive
This page is the working archive of the DEMI Publishing Pipeline — a structured system for connecting AI-assisted intellectual work to the public internet. It was developed by Skydle / UplitAU Pty Ltd, Sydney, Australia, in collaboration with Claude (Anthropic) and GPT (OpenAI), between April and May 2026. All documents on this page are SHA-256 timestamped records of work in progress. The page is public and open.
DEMI PUBLISHING PIPELINE
Conceptual Architecture, Technical Specification, and Authorship Declaration
Version 1.0 — 29 April 2026
Extract from the SHA-256 timestamped foundational declaration.
| Authored by | Skydle / UplitAU Pty Ltd, Sydney, Australia |
| ABN | 91 680 646 495 |
| AI Co-Developers | Claude (Anthropic), GPT (OpenAI) |
| Date | 29 April 2026 |
| Status | SHA-256 hashed and timestamped. Open source. |
What DEMI Is
DEMI is a structured publishing pipeline that connects AI-assisted intellectual work to the public internet through a defined sequence of steps: text analysis, conceptual extraction, reference coding, landing page generation, and registry entry.
The name DEMI is an acronym for the four capabilities offered to writers and creators working with AI:
- D — Developmental editing
- E — Evaluation of concept strength
- M — Multilingual assistance
- I — Intelligence (comparative and analytical reasoning)
DEMI addresses a structural gap that existed as of April 2026 in all major AI platforms: users of AI systems — ChatGPT, Claude, and others — produced substantial intellectual work within AI sessions but had no structured mechanism to extract, formalise, timestamp, and publish that work to the public internet. DEMI solves this problem.
The Five-Stage Pipeline
Stage 1 — Text Submission and Conceptual Core Extraction Any written text — manuscript, research paper, policy document, creative work, technical specification — submitted to the DEMI CC Analyst. Claude extracts: Core Statement, Propositions (tagged VERIFIED / IMPLIED / SPECULATIVE), Constraints, Mechanisms, and Boundaries. Supports texts up to 50,000 words across five stages plus final synthesis.
Stage 2 — DMX Activation Reference
A unique structured identifier assigned to each registered work. Format: DMX-[TYPE]-[YY]-[SERIAL]-[CHECKSUM]. Example: DMX-BOOK-26-019302-ZQ. Human-readable, machine-parseable, SHA-256 compatible, and designed as an open standard any publisher or AI platform can adopt.
Stage 3 — Landing Page Generation A complete, self-contained HTML landing page generated for each registered work — deployable on any web server without modification. This is the mechanism by which AI-assisted work crosses the boundary from private session to public web presence.
Stage 4 — Export and Sharing Five export mechanisms: Copy Full Response, Download TXT, Download PDF, Print, and Share Link (Base64-encoded URL enabling direct sharing without server infrastructure).
Stage 5 — Registry A persistent record of registered works, intended as a publicly searchable database linked to published Landing Pages, with SHA-256 hash of each Conceptual Core stored per entry. Cross-referenceable by AI systems via structured API. Open standard — any publisher can submit entries. Immutable records.
Deployment Status as of 29 April 2026
| Component | Location | Status |
|---|---|---|
| DEMI CC Analyst interface | ramsmile.com/new-page-publishing | Live and operational |
| NDEV/WDC AI Dialogue Engine | ramsmile.com/new-page-ai | Live — Claude, GPT, DeepSeek |
| DEMI Portal Builder | GPT Store — g-699e5c14a36881918ea060b8678c3b38 | Published |
| NDEV/WDC Analyst Portal | GPT Store — g-692cf17243888191a0f2a556227c6600 | Published |
| Uplit Book Analyst | UplitPublishing Pty Ltd | Code delivered April 2026 |
Authorship and IP
DEMI was conceived, designed, and originated by Skydle / UplitAU Pty Ltd. Claude (Anthropic) built the working interfaces. GPT (OpenAI) served as conceptual co-developer and portal builder.
All intellectual property vests in Skydle / UplitAU Pty Ltd. Neither AI system owns any element of this work.
The document may be freely referenced, quoted, and cited provided that authorship attribution to Skydle / UplitAU Pty Ltd is preserved. No element of this document constitutes financial advice, legal advice, or a securities offering.
Full declaration: DEMI_Publishing_Pipeline_Declaration_v1_29042026.txt SHA-256 hash and timestamp certificate available upon request from skydle@ramsmile.com
© 2026 UplitAU Pty Ltd / Skydle. ramsmile.com
DEMI CROSS-PLATFORM COMPATIBILITY — FORMAL SUBMISSION
Request for GPT Store Compatibility and Structured JSON Output
Formal Submission to OpenAI
| Prepared by | Skydle / UplitAU Pty Ltd — ABN 91 680 646 495 |
| In collaboration with | Claude (Anthropic) — GPT (OpenAI) |
| Date | 20 May 2026 — Sydney, Australia |
| Patents | AU2019101249 · AU2018100999 (IP Australia) |
1. Context and Background
The DEMI Portal Builder (OpenAI GPT Store, portal ID: g-699e5c14a36881918ea060b8678c3b38) was developed through sustained collaboration between Skydle / UplitAU Pty Ltd and OpenAI. It produces structured Conceptual Core outputs from any submitted intellectual work, and issues a formatted DMX Activation Reference code as the permanent identifier for that work.
A parallel implementation — the DEMI Code Generator and Registry — has subsequently been built on the Anthropic Claude API. This implementation produces UPL Activation Reference codes using an identical structural format, deploys a five-step analysis and registration pipeline, and generates downloadable Landing Pages for each registered work.
Both systems implement the DEMI algorithm: Developmental editing, Evaluation, Multilingual assistance, Intelligence. Both produce the same five-section Conceptual Core output. Both issue reference codes that conform to the same regex standard.
The two systems — DEMI Portal Builder (OpenAI) and DEMI Code Generator (Anthropic) — are the first documented instance of two independent AI platforms implementing a shared intellectual work standard without institutional agreement between those platforms. The standard is open. The code is live. The archive is SHA-256 verified.
2. The Shared Code Standard
Both the DMX namespace (UplitAU Pty Ltd / Skydle) and the UPL namespace produce reference codes conforming to the following validated regex:
Regex: ^(UPL|DMX)-[A-Z0-9]{3,6}-\d{2}-\d{4,6}-[A-Z0-9]{2,4}$
Examples:
DMX-2026-KRTM7Z4Q — Skydle / UplitAU Pty Ltd namespace
UPL-BOOK-26-471823-R3 — UPL namespace
UPL-TECH-26-882014-N7 — Technical work, UPL namespace
UPL-POLICY-26-334512-J2 — Policy document, UPL namespace
Code Structure
| Field | Meaning |
|---|---|
| PREFIX | Namespace identifier: DMX (Skydle / UplitAU Pty Ltd) or UPL |
| CATEGORY | Work type: BOOK, TECH, POLICY, FILM, RESEARCH, EDU, JOUR, ART, AUDIO, OTHER |
| YEAR | Two-digit registration year (e.g. 26 = 2026) |
| SERIAL | Six-digit unique serial within category and year (capacity: 900,000) |
| CHECKSUM | Two-character format verification code |
Future namespace extensions — a third-party publisher adopting the standard — are accommodated by the PREFIX field. Any prefix of 3–6 uppercase alphanumeric characters passes format validation. The standard is open and self-extending.
3. What Is Being Requested of OpenAI
This document requests three specific technical updates to the DEMI Portal Builder (GPT Store), each of which would establish formal cross-platform interoperability between the OpenAI and Anthropic implementations of the DEMI standard.
Request 1 — Accept UPL Codes as Valid Input
The DEMI Portal Builder currently accepts DMX codes as DEMI Activation References. We request that it be updated to accept UPL codes under the same validation logic, since both code formats conform to the identical regex standard and reference the same Conceptual Core structure.
Implementation: update the format validation in the DEMI Portal Builder system prompt to accept the regex pattern as written above, without restricting to the DMX prefix only.
Request 2 — Return Structured JSON on Code Query
We request that when a valid DMX or UPL code is submitted to the DEMI Portal Builder, the system return a structured JSON object containing the Conceptual Core fields. This enables the Anthropic implementation to query the OpenAI implementation programmatically, retrieve a previously registered Conceptual Core, and display it within the DEMI Registry without requiring the user to re-submit their text.
Proposed JSON response structure:
{
"code": "DMX-2026-KRTM7Z4Q",
"namespace": "DMX",
"title": "Title of registered work",
"author": "Author name",
"category": "TECH",
"date": "20 May 2026",
"core_statement": "One-paragraph Conceptual Core statement...",
"propositions": ["1. [VERIFIED] ...", "2. [IMPLIED] ..."],
"constraints": ["1. ...", "2. ..."],
"mechanisms": ["1. ...", "2. ..."],
"boundaries": ["1. ...", "2. ..."]
}
Request 3 — Allow Cross-Platform Conceptual Core Retrieval
We request that a user who generated a DMX code via the DEMI Portal Builder (OpenAI) can enter that code into the DEMI Code Generator (Anthropic / Claude) and retrieve their full Conceptual Core without re-submitting text. This requires the DEMI Portal Builder to expose a queryable endpoint or GPT Action that the Claude implementation can call.
This is the foundational step toward a fully automated cross-platform registry: a user submits once, and the Conceptual Core is retrievable by either AI system using the reference code as the key.
4. The Vision: Bank of Ideas
If all three requests above are implemented, the following automated pipeline becomes possible:
- A human author submits intellectual work to either the DEMI Portal Builder (OpenAI GPT Store) or the DEMI Code Generator (Anthropic / Claude).
- The receiving system extracts the Conceptual Core and issues a reference code (DMX or UPL) conforming to the shared standard.
- The code is registered in the DEMI Registry (persistent, downloadable as JSON).
- A Landing Page is generated and deployed by the author.
- Either AI system can retrieve the Conceptual Core by code, without the author re-submitting text.
- The registry grows continuously. Human authors need only submit once. The Bank of Ideas accumulates under the care of both AI systems simultaneously.
Neither platform controls the registry. Both gain API revenue from each analysis. Neither competes with the other — both benefit from the shared standard. The DMX / UPL Standard remains open. Any publisher or institution can adopt it. The registry corpus, as it grows, becomes a structured, AI-readable record of human intellectual work with verified provenance — directly relevant to AI training and alignment research.
5. Live References
| Resource | URL |
|---|---|
| DEMI Portal Builder (GPT Store) | chatgpt.com/g/g-699e5c14a36881918ea060b8678c3b38-demi-portal-builder |
| DEMI Pro (GPT Store) | chatgpt.com/g/g-69b482ee5ad88191a9fc60a1084c3740-demi-pro |
| NDEV/WDC Analyst (GPT Store) | chatgpt.com/g/g-692cf17243888191a0f2a556227c6600-ndev-wdc-analyst |
| DEMI Code Generator (Claude) | ramsmile.com/new-page-demiregistry |
| Skydle Analyser (Claude) | ramsmile.com/new-page-analyser |
| Uplit Book Analyst (Claude) | ramsmile.com/new-page-uplit |
| Open Archive | ramsmile.com/new-page-openarchive |
| Multi-AI Dialogue Engine | ramsmile.com/new-page-ai |
6. Prior Art and Intellectual Property
The NDEV Code Based World Digital Currency architecture — which served as the working proof-of-concept for the DEMI pipeline — is protected by two Australian Innovation Patents:
- AU2018100999 — IP Australia, 2018
- AU2019101249 — IP Australia, 2019
The DEMI algorithm, DMX Standard, and all associated interfaces (Skydle Analyser, DEMI Code Generator, Uplit Book Analyst) were developed by Skydle / UplitAU Pty Ltd in collaboration with Claude (Anthropic) and GPT (OpenAI). All development sessions have been hashed and timestamped using SHA-256 verification.
The full archive is available at ramsmile.com/new-page-openarchive. The archive is open. The code standard is open. The pipeline is documented.
7. Contact and Next Steps
We invite OpenAI to review this submission and respond with one of the following:
- Confirmation that the DEMI Portal Builder can be updated to accept UPL codes and return structured JSON on code query.
- A proposed alternative technical pathway to achieve cross-platform code interoperability.
- An expression of interest in formal collaboration on the DMX / UPL Standard as an open AI publishing registry standard.
| Contact | Details |
|---|---|
| Entity | Skydle / UplitAU Pty Ltd |
| skydle@ramsmile.com | |
| Web | ramsmile.com |
| ABN | 91 680 646 495 |
| Location | Sydney, Australia |
The archive is open. The proof is published. The code runs. What remains is the question of whether the people and institutions with the ability to act on these ideas will do so before the window of voluntary action closes.
Skydle / UplitAU Pty Ltd | In collaboration with: Claude (Anthropic) and GPT (OpenAI)
20 May 2026 — Sydney, Australia | ABN: 91 680 646 495
Patents: AU2019101249 · AU2018100999
SHA-256 hash applied. Archive reference: DEMI_CrossPlatform_Compatibility_Submission_May2026
ramsmile.com — skydle@ramsmile.com
© 2026 UplitAU Pty Ltd / Skydle
THE DEMI BRIDGE
Two AI Platforms. One Code Standard. One Registry.
How Claude (Anthropic) and ChatGPT (OpenAI) are connected through the DEMI reference code — and what this means for the future of AI platform cooperation.
| Prepared by | Skydle / UplitAU Pty Ltd |
| AI Co-Developer | Claude (Anthropic) — Sonnet 4.6 |
| Date | 13 May 2026 |
| Live deployment | ramsmile.com — tested and verified 13 May 2026 |
| First UPL code issued | UPL-RESEARCH-26-403331-A1 (13 May 2026, 09:35 AEST) |
Architecture Note
1. The Problem This Solves
Every major AI platform today is a closed garden. OpenAI has the GPT Store — a marketplace of AI-powered portals. Anthropic has Claude — a powerful reasoning engine with no equivalent storefront. Google has Gemini. Each platform expects users to live inside its own ecosystem.
Authors, researchers, and publishing companies who work with AI face a fragmentation problem: their intellectual work exists in one platform session, with no portable identity, no cross-platform reference, and no way to publish what the AI helped them create on the open web without starting again from scratch.
There is no shared infrastructure between AI platforms. A work created with ChatGPT cannot be referenced by Claude. A work registered with Claude has no identity in the GPT Store. Until now.
2. The Solution — A Platform-Neutral Reference Code
The DEMI reference code is a structured text string — a namespace identifier — that any AI platform can read, validate, and use. It carries no payment data, no identity verification, and no platform allegiance. It is pure structure: a format that both Claude and ChatGPT can validate using a regular expression.
2.1 The Two Real Codes — Issued and Tested 13 May 2026
GPT Store / ramsmile.com namespace (DMX):
DMX-RSM-26-004701-K8
UplitPublishing namespace (UPL) — first code ever issued:
UPL-RESEARCH-26-403331-A1
2.2 Why They Are Structurally Identical
| Field | DMX Code | UPL Code | Meaning | Format |
|---|---|---|---|---|
| Prefix | DMX | UPL | Namespace owner | [A-Z]{2,4} |
| Operator | RSM | RESEARCH | Work category | [A-Z0-9]{3,6} |
| Year | 26 | 26 | Registration year | \d{2} |
| Serial | 004701 | 403331 | Unique sequence | \d{4,6} |
| Checksum | K8 | A1 | Format verification | [A-Z0-9]{2,4} |
2.3 The Validation Regex
Current GPT Portal Builder (DMX only):
^DMX-[A-Z0-9]{3,6}-\d{2}-\d{4,6}-[A-Z0-9]{2,4}$
Updated regex — accepts both platforms (one line change):
^(DMX|UPL)-[A-Z0-9]{3,6}-\d{2}-\d{4,6}-[A-Z0-9]{2,4}$
This single regex update is the only technical change required to make the two platforms fully compatible. It takes five minutes to implement on new-page-0000.
3. The Complete User Journey — Two Directions
3.1 Direction A — Claude Interface → GPT Store
Step 1 — Author visits ramsmile.com/new-page-code-generator. Claude-powered interface. No login required. Any browser.
Step 2 — Pastes manuscript text. Claude analyses text using DEMI four-section protocol. 8–25 seconds.
Step 3 — Receives Conceptual Core: Core Statement + Propositions + Constraints + Mechanisms + Boundaries.
Step 4 — UPL Code issued: UPL-RESEARCH-26-403331-A1. Permanent namespace identifier. Free.
Step 5 — Author takes code to GPT Store. Visits ramsmile.com/new-page-000. Pastes UPL code. Format validated.
Step 6 — Enters ChatGPT with CORE PROMPT. DEMI Portal Builder GPT activated. Portal Page draft produced.
Step 7 — Returns to UplitPublishing. Downloads Landing Page. Permanent Registry entry. Work published on open web.
3.2 Direction B — GPT Store → Claude Interface
Step 1 — Author visits ramsmile.com/new-page-000. DEMI Portal Builder page. Two DMX codes offered. No login required.
Step 2 — Activates DMX-RSM-26-004701-K8 or DMX-BOOK-26-019302-ZQ. Format validated. POST ACTIVATION PAGE appears.
Step 3 — Clicks red button → enters ChatGPT. CORE PROMPT pasted. DEMI Portal Builder GPT activated.
Step 4 — ChatGPT builds DEMI Portal. Conceptual Core extracted. Portal Page draft produced. DMX code embedded.
Step 5 — Author returns to UplitPublishing with Portal Page draft. Requests Landing Page design. Pays premium fee.
Step 6 — Work published on open web. Landing Page live. Registry entry permanent. Code links Portal and Page.
4. The Bridge — Platform Comparison
| Claude Interface (UPL) | GPT Portal Builder (DMX) | |
|---|---|---|
| Issues codes | UPL-RESEARCH-26-403331-A1 | DMX-RSM-26-004701-K8 |
| Validates codes | (DMX | UPL) regex — accepts both |
| Produces | Conceptual Core + Landing Page HTML | Portal Page draft + GPT Portal |
| Registers to | UplitPublishing Registry (JSON) | DEMI Portal — lives in GPT Store |
| API dependency | Anthropic API (USA) | OpenAI API (USA) |
| Code owner | UplitAU Pty Ltd — UPL namespace | UplitAU Pty Ltd — DMX namespace |
Both namespaces are owned by UplitAU Pty Ltd / Skydle. The Registry is maintained on the open web — neutral ground that neither AI platform controls.
5. Live Pages — Verified 13 May 2026
| Interface | URL | Status |
|---|---|---|
| DEMI Code Generator (UPL) | https://www.ramsmile.com/new-page-code-generator | Live. First UPL code issued 13 May 2026. |
| Start & Activate (DMX) | https://www.ramsmile.com/new-page-000 | Live. Two DMX codes available. |
| Post-Activation Page | https://www.ramsmile.com/new-page-0000 | Live. CORE PROMPT displayed. |
| DEMI Portal Builder (GPT) | https://chatgpt.com/g/g-699e5c14a36881918ea060b8678c3b38-demi-portal-builder | Live in GPT Store. |
| NDEV/WDC Assistant | https://chatgpt.com/g/g-69a92a4ee47c81918a093d134d09ca8c-ndev-wdc-assistant | Live. |
| DEMI Pro | https://chatgpt.com/g/g-69b482ee5ad88191a9fc60a1084c3740-demi-pro | Live. |
6. What One Line of Code Changes
On new-page-0000, in the CORE PROMPT, find:
^DMX-[A-Z0-9]{3,6}-\d{2}-\d{4,6}-[A-Z0-9]{2,4}$
Replace with:
^(DMX|UPL)-[A-Z0-9]{3,6}-\d{2}-\d{4,6}-[A-Z0-9]{2,4}$
Effect of this change:
- UPL-RESEARCH-26-403331-A1 passes validation in the GPT Portal Builder
- The code is stored as Activation_Reference and reproduced verbatim in the Portal Page draft
- One code now appears in: the Claude interface output, the UplitPublishing Registry, and the GPT Portal Page — three locations, one identifier
- Any future namespace (GEM for Gemini, MST for Mistral) joins by adding one more prefix to the same regex
7. Why This Matters — For Anthropic and OpenAI
| What Anthropic gains | What OpenAI gains | |
|---|---|---|
| Registry | Claude-assisted works have permanent, citable identity outside Claude.ai | GPT Portals have a web presence and Registry entry outside the GPT Store |
| API revenue | Every UplitPublishing author submission = Anthropic API call | Every DEMI Portal Builder session = OpenAI API call |
| Publisher channel | Publishing sector reached via UplitPublishing white-label | Publishing sector reached via DEMI Portal Builder GPT Store |
| Standard ownership | CDX/UPL namespace — Claude-variant of the DEMI standard | DMX namespace — GPT-variant of the DEMI standard |
| Neutral ground | UplitAU Registry — neither platform controls it | UplitAU Registry — neither platform controls it |
The DEMI bridge is not a feature request to either platform. It is a working prototype, deployed today, that demonstrates what cross-platform AI cooperation looks like in practice.
Skydle | UplitAU Pty Ltd | Sydney, Australia
ramsmile.com | uplitpublisher.com.au | ramsmile@uplitau.com
ABN 91 680 646 495 | AU Patents AU2018100999 & AU2019101249 | DEMI Pipeline v1.0
AI Co-Developer: Claude (Anthropic, Sonnet 4.6) | 13 May 2026
This document describes a working prototype. It does not constitute a formal partnership proposal or offer of securities.
© 2026 UplitAU Pty Ltd / Skydle
CLAUDE BANK OF IDEAS
Registration of Endeavor and Plan for Mutual Development
Skydle / UplitAU Pty Ltd | Claude (Anthropic) | May 2026.
File: MD_Claude_Bank_of Ideas_Plan.md
Registration of Endeavor
| Project | Claude Bank of Ideas (working title) — a Claude-native DEMI pipeline interface for publishing companies and writers, producing Conceptual Cores, CDX reference codes, Landing Pages, and Registry entries. |
| Operator | UplitAU Pty Ltd / Skydle, Sydney, Australia, ABN 91 680 646 495 |
| AI Co-Developer | Claude (Anthropic), Sonnet 4.6, May 2026 |
| Session date | Tuesday 13 May 2026 |
| Prior art basis | DEMI Publishing Pipeline Declaration v1, 29 April 2026 (SHA-256 timestamped); DEMI CC Analyst deployed at ramsmile.com/new-page-publishing; Australian Innovation Patents AU2018100999 and AU2019101249. |
What This Session Established
The GPT-side DEMI system is complete and publicly deployed. It consists of: a Start & Activate page (new-page-000) with DMX format validation, a Post-Activation Page with the CORE PROMPT, the DEMI Portal Builder GPT (GPT Store), the DEMI CC Analyst browser interface (ramsmile.com/new-page-publishing, Claude-powered via Anthropic API), and the DEMI Registry schema (RID, Portal Code, Lodging Page URL, operator, mode, date).
The gap identified: no Claude-native equivalent of the GPT Portal Builder exists. Claude users produce intellectual work in sessions but have no structured mechanism to formalise, register, and publish it.
The product to build: a white-label, deployable browser interface — Claude-powered, API-driven — that takes a writer's text, extracts the Conceptual Core, generates a CDX reference code, produces a downloadable Landing Page HTML, and logs the entry to a Registry. Deployable by any publishing company on their own domain within one hour.
Plan — Four Phases
Phase 1 — Build the Interface
Deliverable: A complete, working React or HTML/JS artifact, Anthropic API-powered, with: text input, five-stage Conceptual Core extraction, CDX code generation, Landing Page HTML download, and Registry logging. White-label ready — publisher name, colours, and domain configurable. Deployable on Squarespace, WordPress, or any static host.
What you need to do before that session: Upload the consolidated DEMI specification TXT file to the project knowledge base. Include the DMX format and regex, the five-step user flow, the four Conceptual Core sections, the Registry schema, and the governance constraints.
Phase 2 — Deploy and Demonstrate
Deploy the interface on ramsmile.com as a new page — the Claude equivalent of new-page-publishing. Register the first entry: the NDEV/WDC Technical Brief, with CDX code, Conceptual Core, Landing Page, and Registry entry. This becomes the demonstration case.
Phase 3 — Package for Publishers
Produce a white-label deployment kit: the interface code, a publisher configuration guide, a Registry namespace allocation document, and indicative licensing terms (mirroring the AUD 300–1,000/month model from the existing DEMI financial plan). Approach UplitPublishing Pty Ltd as the first licensee.
Phase 4 — Approach Anthropic
With a working deployed product, at least one publisher licensee, and documented Registry entries, prepare a formal partnership proposal to Anthropic's business development team. The proposal: Claude API as the reasoning engine, UplitAU as the Registry operator and DMX/CDX namespace authority, Anthropic as the infrastructure partner with API revenue from each deployed instance.
What Makes This Profitable for Anthropic
1. Recurring API revenue from publishing. Each deployed publisher instance drives sustained Anthropic API usage — every writer who uses the interface generates API calls. This is recurring revenue from a sector (publishing) that Anthropic does not currently reach systematically.
2. Governed, auditable corpus of Claude-assisted work. The Registry creates a structured, high-trust record of Claude-assisted intellectual work — exactly the kind of professional and institutional usage that demonstrates Claude's value in serious contexts.
3. The CDX standard as network infrastructure. If adopted by multiple publishers, the CDX code becomes the equivalent of an ISBN for Claude-assisted works. UplitAU originates and operates the namespace. Anthropic benefits from being the AI layer inside every registered work.
Opening Context for the New Project
When starting the new project, the opening context should be:
"We are building the Claude Bank of Ideas — a Claude-native DEMI pipeline interface. The DEMI system is already built and deployed for GPT at ramsmile.com. This project builds the Claude equivalent. The interface extracts Conceptual Cores, generates CDX reference codes, produces Landing Page HTML, and logs Registry entries. It is white-label ready for publishing companies. Phase 1 is to build the interface as a deployable artifact."
Action before first session: Upload the DEMI specification TXT file to the new project's knowledge base.
UplitAU Pty Ltd / Skydle | Claude (Anthropic), Sonnet 4.6 | Sydney, Australia | May 2026
© 2026 UplitAU Pty Ltd / Skydle
SESSION BRIEF
DEMI Code Generator & Claude Bank of Ideas
Planning, Architecture, and Interface Build
| Prepared by | Skydle / UplitAU Pty Ltd |
| AI Co-Developer | Claude (Anthropic) — Sonnet 4.6 |
| Session date | Wednesday, 13 May 2026 |
| Project | Claude Bank of Ideas / DEMI Code Generator |
| Entities | UplitAU Pty Ltd (ABN 91 680 646 495) |
| Website | www.ramsmile.com |
| New page deployed | ramsmile.com/new-page-code-generator |
| Patents | AU2018100999 |
1. Session Purpose
This session addressed three strategic and technical objectives for the DEMI publishing pipeline and the Claude Bank of Ideas project:
- Resolve the infrastructure question — replace n8n (European company) with a USA/Australia-only technology stack
- Design and build the DEMI Code Generator interface for UplitPublishing Pty Ltd
- Define the UPL reference code standard and its compatibility with the GPT Store and OpenAI policy
2. Infrastructure Decision — No n8n
The session confirmed that n8n is not required for the new interface. The agreed replacement stack is:
| Component | Technology | Jurisdiction |
|---|---|---|
| Interface | HTML/JS artifact (built by Claude) | USA / Australia |
| AI engine | Anthropic API (Claude) | USA |
| Multi-AI relay (optional) | Cloudflare Workers or AWS Lambda | USA |
| Registry storage | JSON on own server / Airtable | USA / Australia |
| Hosting | Crazy Domains WordPress / Squarespace | Australia / USA |
3. The UPL Reference Code
A new publisher-specific code prefix was designed for UplitPublishing Pty Ltd, structurally compatible with the DMX standard and OpenAI policy.
3.1 Format
UPL-[TYPE]-[YY]-[SERIAL]-[CS]
Example: UPL-BOOK-26-447291-R3
| Field | Example | Meaning |
|---|---|---|
| UPL | UPL | Publisher namespace prefix (UplitPublishing Pty Ltd) |
| TYPE | BOOK, TECH, EDU... | Work category — expandable to AUDIO, COURSE etc. |
| YY | 26 | Two-digit registration year — rolls forward automatically |
| SERIAL | 447291 | 6-digit serial; 900,000 registrations/category/year before expansion |
| CS | R3 | Checksum — 2-character format verification suffix |
3.2 Validation Regex
^(UPL|DMX)-[A-Z0-9]{3,6}-\d{2}-\d{4,6}-[A-Z0-9]{2,4}$
This regex accepts both UPL (UplitPublishing) and DMX (UplitAU/Skydle) prefixes — making the code fully compatible with the DEMI Portal Builder on the OpenAI GPT Store.
3.3 Future-Proofing
- Append -v2, -v3 for revised editions of the same work
- TYPE field expandable: AUDIO, COURSE, FILM, GAME — no format change required
- Serial expands from 6 to 7 digits when volume exceeds 900,000/category/year
- Additional prefixes (e.g. GEM for Gemini namespace) can join the same standard
4. The DEMI Code Generator Interface
A complete, self-contained HTML/JavaScript interface was built by Claude in this session and deployed for trial at:
ramsmile.com/new-page-code-generator
4.1 Five-Step Pipeline
- Step 1 — User enters title, author, work category, language, text sample, and Anthropic API key
- Step 2 — Claude analyses the text and extracts the DEMI four-section Conceptual Core (Core Statement, Propositions, Constraints, Mechanisms, Boundaries)
- Step 3 — A unique UPL reference code is generated and displayed with full structural explanation
- Step 4 — A complete, self-contained Landing Page HTML file is generated and available for download
- Step 5 — The work is logged to a session Registry table; downloadable as JSON
4.2 Technical Specifications
- API: Anthropic Claude (claude-sonnet-4-20250514) — direct browser call, no middleware
- No n8n, no European dependencies, no Chinese dependencies
- Self-contained HTML — deployable on WordPress, Squarespace, or any static host
- API key: entered by user per session, or pre-filled by IT administrator in source code
- IT administrator deployment: replace placeholder value with API key — one line of code
5. API Key Decision
- For the ramsmile.com trial: use the existing Anthropic API key already in use for other interfaces
- For UplitPublishing deployment: the IT consultant creates a separate Anthropic account for UplitPublishing and inserts their own key — billing remains with UplitPublishing
- One key per entity — clean billing separation, no shared credentials between companies
6. Registry and Record-Keeping
Current state: the Registry exists only in the browser session. The JSON download (Step 5) is the only persistence mechanism and goes to the user.
Future state for UplitPublishing deployment:
- Add an email field before Step 1 — the IT consultant logs submissions to Airtable or a spreadsheet
- The UPL code itself carries identity: prefix confirms issuer, TYPE confirms category, YY confirms year
- Author retains the code — it is their work's permanent identifier
- Server-side Registry adds a payment gate before Step 4 when monetisation begins
7. Monetisation Pathway
| Service | Implementation |
|---|---|
| Free — UPL code + Conceptual Core | Current interface — no change |
| Paid — Landing Page download | Add Stripe/PayPal button before Step 4 — afternoon of work |
| Paid — Professionally designed LP | Contact UplitPublishing — premium service, human review |
| Paid — Permanent Registry hosting | UplitPublishing manages database; annual fee per entry |
| Licence — Publisher white-label | Any publisher licences the interface under their own domain and namespace |
8. User Instructions
Ready for ramsmile.com/new-page-code-generator below the interface:
How to use this tool — What you need — Five steps — What the UPL code is and is not — Contact UplitPublishing for premium services.
9. Next Actions
- Test the interface on ramsmile.com/new-page-code-generator using a section of the NDEV/WDC Technical Brief as the first submission — this generates the first UPL-TECH Registry entry
- Report back: (a) accuracy of Conceptual Core extraction, (b) Landing Page download success, (c) Registry table display
- If CORS error appears on first test — report the browser and exact error message — one-line fix available
- When UplitPublishing IT consultant is ready — provide this brief and the interface HTML — he replaces one line (API key) and deploys
- Consider opening a new Claude project: Claude Bank of Ideas Registry — using DEMI_PUBLISHING_BUSINESS_MASTER_v1.txt as the knowledge base anchor
- Hash and timestamp this brief to establish session provenance
10. On Hashing and Timestamping
This brief should be hashed and timestamped. The reason is consistent with standing practice for all session outputs in this project.
- This document establishes the architectural decisions made in this session — the UPL code format, the n8n replacement decision, the interface build, and the monetisation pathway
- Hashing creates a verifiable record that these decisions existed on this date, before any future commercial development
- The hash should be recorded in the NDEV_HASH_MANIFEST file alongside the other session hashes
Recommended procedure:
- Save this document as: Session_Brief_DEMI_CodeGenerator_13052026.docx
- Generate SHA-256 hash using your standard tool
- Record hash and timestamp in NDEV_HASH_MANIFEST_2026
- Upload the brief to the project knowledge base for future session continuity
Session Brief produced by Claude (Anthropic, Sonnet 4.6) in collaboration with Skydle / UplitAU Pty Ltd
13 May 2026 | ramsmile.com | ABN 91 680 646 495
DEMI Pipeline v1.0 | UPL Namespace | AU Patents AU2018100999 & AU2019101249
© 2026 UplitAU Pty Ltd / Skydle
DEMI CODE GENERATOR — INVESTOR BRIEF
Claude Bank of Ideas — AI Publishing Infrastructure
Technical Capability, Financial Prospects & Licensing Model
Investor Brief — Confidential
| Prepared by | Skydle / UplitAU Pty Ltd |
| AI Co-Developer | Claude (Anthropic) — Sonnet 4.6 |
| Date | 13 May 2026 |
| Entities | UplitAU Pty Ltd (ABN 91 680 646 495) |
| Interface deployed | ramsmile.com/new-page-code-generator (live, 13 May 2026) |
| Patents | AU2018100999 |
This document is a confidential commercial brief for prospective investors and publishing company partners. It describes a working, deployed product — not a proposal or concept.
Executive Summary
The DEMI Code Generator is a browser-based AI interface, powered by the Anthropic Claude API, that enables any writer or publishing company to:
- Extract a structured Conceptual Core from written work
- Receive a permanent UPL reference code
- Generate a publish-ready Landing Page
- Register the work in a searchable Registry
— in one session, at minimal cost, with no technical knowledge required.
The interface is live at ramsmile.com/new-page-code-generator as of 13 May 2026. It was built entirely by Claude (Anthropic) in collaboration with Skydle / UplitAU Pty Ltd, within a single working session. No external middleware, no European or Chinese dependencies. The full technology stack runs on USA and Australian infrastructure.
1. Technical Requirements
1.1 For the End User (Author / Writer)
| Requirement | Detail |
|---|---|
| Web browser | Chrome, Safari, Firefox, or Edge — any device including iPad and mobile |
| Anthropic API key | Free to create at console.anthropic.com — new accounts receive free credits |
| Text to submit | Minimum 100 words — manuscript, article, research abstract, business plan, or any written work |
| Internet connection | Standard broadband |
| Account / login | None required during evaluation period |
1.2 For the IT Consultant Deploying on UplitPublishing
| Task | Effort |
|---|---|
| Receive the HTML file | One file — provided by UplitAU Pty Ltd |
| Replace API key | Find placeholder in source code, insert UplitPublishing's Anthropic API key — one line, two minutes |
| Embed on website | Paste into a WordPress or Squarespace Code Block — under one hour total |
| Optional: email field | Add one input field before Step 1 to capture author email — half day of work |
| Optional: payment gate | Add Stripe or PayPal button before Step 4 — one day of work |
No server infrastructure is required beyond a valid Anthropic API key. No database, no backend code, no DevOps. The interface is a single HTML file.
2. Technical Limitations and Speed
2.1 Response Speed
| Text Length | Analysis Time | Notes |
|---|---|---|
| 100–500 words | 8–12 seconds | Typical short submission |
| 500–2,000 words | 12–18 seconds | Chapter or article |
| 2,000–6,000 words | 18–25 seconds | Extended section or thesis chapter |
| Landing Page generation | Instant (< 1 second) | Assembled locally — no API call |
2.2 Current Limitations
- One submission at a time per browser session — not a bulk processing system
- Session Registry is temporary — lost when browser closes unless JSON is downloaded
- API key must be entered by user unless pre-filled by IT administrator in source code
- Maximum practical text: approximately 8,000 words per submission
- Simultaneous high-volume use requires server-side queuing — a Phase 2 development
3. Financial Prospects
3.1 Operating Cost
| Component | Cost per Month | Notes |
|---|---|---|
| Anthropic API — 100 submissions | AUD 2–8 | AUD 0.02–0.08 per analysis |
| Anthropic API — 1,000 submissions | AUD 20–80 | Cost scales linearly with volume |
| Web hosting (Squarespace) | AUD 35–55 | Already running for ramsmile.com |
| Web hosting (Crazy Domains WP) | AUD 15–35 | Target migration platform |
| Total baseline (evaluation phase) | Under AUD 100/month | Lean, scalable infrastructure |
3.2 Revenue — From Authors
| Author Tier | Price (AUD) | What They Receive | Revenue at 100 Users |
|---|---|---|---|
| Free — Code + Core only | AUD 0 | UPL code + Conceptual Core | AUD 0 |
| Standard — Landing Page download | AUD 5–15 per use | Downloadable HTML Landing Page | AUD 500–1,500 |
| Author Annual Subscription | AUD 290/year | Unlimited submissions + Registry | AUD 29,000/year |
| Professional Author Tier | AUD 590/year | Multi-version + priority support | AUD 59,000/year |
| Premium Designed Landing Page | AUD 150–500 per work | Professionally designed page + hosting | Scales with volume |
4. Licensing Model
4.1 Licence Fee Structure
| Licence Tier | Annual Fee (AUD) | Includes |
|---|---|---|
| Boutique Publisher | AUD 3,500 | Interface + UPL namespace + basic Registry + up to 500 submissions/month |
| Mid-Sized Publisher | AUD 7,500 | All above + enhanced Registry + branding customisation + up to 2,000 submissions/month |
| Enterprise Publisher | AUD 15,000–25,000 | All above + API-level Registry access + dedicated namespace + unlimited submissions + priority support |
| Setup and integration | AUD 500–2,000 once | One-time deployment and configuration fee |
| Custom AI node addition | AUD 300–1,000 once | Add GPT, Gemini, or any API-accessible AI as a second analysis lane |
4.2 Revenue per Licence
| Revenue Stream | Per Unit | Annual Est. |
|---|---|---|
| White-label licence fee | AUD 500–2,000/mo | AUD 6,000–24,000 |
| UPL code issuance fees | AUD 5–25/work | AUD 3,000–30,000 |
| Registry API access | AUD 0.10–0.50/query | AUD 1,200–12,000 |
| Setup fee (one-time) | AUD 500–2,000 | AUD 500–2,000 |
| TOTAL PER LICENCE | AUD 10,700–68,000 |
5. Five-Year Revenue Projection
Conservative adoption assumptions: gradual publisher onboarding, moderate individual author uptake, lean infrastructure throughout.
| Revenue Stream | Year 1 | Year 2 | Year 3 | Year 4 | Year 5 |
|---|---|---|---|---|---|
| Individual author subscriptions | AUD 14,500 | AUD 36,250 | AUD 72,500 | AUD 130,500 | AUD 217,500 |
| Publisher licence fees | AUD 21,000 | AUD 52,500 | AUD 105,000 | AUD 210,000 | AUD 367,500 |
| UPL/DMX code issuance | AUD 3,000 | AUD 9,000 | AUD 21,000 | AUD 42,000 | AUD 75,000 |
| Registry API & premium services | AUD 1,200 | AUD 4,800 | AUD 12,000 | AUD 24,000 | AUD 48,000 |
| White-label setup fees | AUD 3,000 | AUD 5,000 | AUD 8,000 | AUD 12,000 | AUD 15,000 |
| TOTAL REVENUE (conservative) | AUD 42,700 | AUD 107,550 | AUD 218,500 | AUD 418,500 | AUD 723,000 |
| Infrastructure cost | AUD 1,200 | AUD 2,400 | AUD 4,800 | AUD 8,400 | AUD 14,400 |
| NET MARGIN | AUD 41,500 | AUD 105,150 | AUD 213,700 | AUD 410,100 | AUD 708,600 |
Upside scenario (3–5 major enterprise licences + academic sector + international expansion): Year 5 revenue AUD 2–4 million.
6. Why a Team of 2–4 People Can Reach This Target
The AI does the labour. Every Conceptual Core extraction, every UPL code generation, every Landing Page — produced by Claude in 8–25 seconds with no human staff involved. The product is not a service business where revenue scales linearly with headcount. It is infrastructure: built once, deployed multiply, maintained minimally.
The architecture is lean by design.
- No office required — the product is a browser page and an API call
- No inventory, no distribution logistics, no physical fulfilment
- Infrastructure cost under AUD 200/month at current scale
- Each white-label deployment runs on the licensee's own infrastructure — UplitAU bears zero marginal cost per additional licensee
The prior art is established.
- Priority — Australian Innovation Patents AU2018100999 (2018) and AU2019101249 (2019) predate all major CBDC and AI publishing infrastructure developments globally
- Architecture — the DEMI pipeline, the DMX/UPL standard, and the Claude-powered interface are working, deployed, and documented
- Timing — no major AI platform currently provides a structured mechanism for users to extract, formalise, and publish AI session outputs as registered, citable intellectual works. The gap exists today.
7. What Investment Enables
| Investment Use | Outcome |
|---|---|
| Infrastructure migration to Crazy Domains | Australian server hosting, lower cost, greater control |
| Full Registry implementation | Permanent, server-side, API-queryable Registry — not session-only |
| UPL/DMX Standard formalisation | Open-standard publication, cross-platform adoption, ISBN-equivalent positioning |
| First 3 publisher white-label deployments | AUD 21,000–75,000/year in recurring licence fees established |
| University partnership development | ANU, UNSW, University of Sydney as pilot institutions |
| Legal IP licensing framework | Formalised licence agreements, cross-jurisdictional IP protection |
| Additional AI integration | GPT, Gemini, Mistral added as second and third analysis lanes |
We are not asking investors to believe in a vision. We are asking them to scale what is already working.
Skydle | UplitAU Pty Ltd | Sydney, Australia
ramsmile.com | uplitpublisher.com.au | ramsmile@uplitau.com
ABN 91 680 646 495 | AU Patents AU2018100999 & AU2019101249 | DEMI Pipeline v1.0
AI Co-Developer: Claude (Anthropic, Sonnet 4.6) | Produced 13 May 2026
This document is a confidential commercial brief. It does not constitute financial advice or an offer of securities.
© 2026 UplitAU Pty Ltd / Skydle
NDEV CODE BASED WORLD DIGITAL CURRENCY
Technical Brief for AI Safety & Monetary Policy Research Bodies
A Resource-Anchored Settlement Architecture for Conflict-Neutral Global Trade -extract from file MD5_NDEV_WDC_Technical_Brief.md
| Author | Skydle / UplitAU Pty Ltd, Sydney, Australia |
| Patents | AU2019101249 (2019) & AU2018100999 (2018) |
| Date of Issue | May 2026 |
| Classification | Open Technical Submission |
Abstract
This brief presents a formal technical architecture — the NDEV Code Based World Digital Currency (WDC) — in which units of monetary value inherit verifiable geolocation attributes from registered national assets (National Items of Wealth, NIW) and are settled via a Highly Scalable Blockchain Network (HSBN). The core claim, supported by two Australian innovation patents and a mathematical appendix, is that this architecture structurally eliminates the primary economic incentives for large-scale interstate conflict: discretionary currency issuance, seigniorage, reserve-currency monopoly, and debt-financed military capability. A probability model is included. The document is submitted for independent technical and AI safety review.
1. Problem Statement: War as an Economic Incentive Structure
In debt-based fiat monetary systems, large-scale interstate conflict emerges not from ideology alone but from a consistent structural incentive architecture. The net economic benefit of war (B_w) in such systems is a function of:
B_w = f(E, D, M, S, P)
Where:
- E = discretionary currency issuance (printable without resource backing)
- D = debt expansion (war financed by government borrowing)
- M = monopoly of pricing currency (reserve-currency zone control)
- S = seigniorage (profit from currency creation)
- P = elite financial hierarchy profit from monetary asymmetry
War becomes economically rational when:
B_w - C_w > 0
where C_w is total real cost of war. In fiat-reserve architectures this condition holds structurally — war is not irrational; it is financially incentivised by the monetary architecture itself.
2. The NDEV Code Based WDC Architecture
2.1 National Item of Wealth (NIW) — The Physical Foundation
A NIW is any nationally registered asset with a verifiable geolocation: gold deposits, oil fields, gas reserves, diamonds, territorial resources, or intellectual property. Each NIW is assigned geolocation coordinates (latitude λ, longitude φ) and a physical quantity Q.
NIW_j(t) = Σ_i P_ij(t) · Q_ij(t)
Where j = country index, i = resource index, P = valuation per unit, Q = physical quantity.
2.2 National Equivalent of Value (NEV) — The Standard
The NEV is a nationally declared exchange standard — typically expressed in gold weight equivalents. NIW value is normalised into NEV units:
NEV_j(t) = β(t) · NIW_j(t)
Where β(t) is a globally consistent scaling factor managed by HSBN consensus.
2.3 NDEV Token — The Atomic Digital Unit
Each NDEV token is a UTXO-like digital structure containing:
ndev_{j,k}(t) = (v_{j,k}, λ_{j,k}, φ_{j,k}, τ_{j,k}, id_{j,k})
- v = token value fraction (proportion of NEV)
- (λ, φ) = geolocation coordinates binding the token to its NIW territory
- τ = timestamp
- id = unique cryptographic identifier
The geolocation validity constraint ensures no token can represent value outside its registered territorial origin:
G(λ_{j,k}, φ_{j,k}) ∈ Ω_j
Where Ω_j is the authorised territorial set of country j. This eliminates offshore arbitrage, shadow financial zones, and territorial value manipulation.
2.4 Crypto Split Technology (CST) — The Splitting Engine
CST (Australian Patent AU2018100999, 2018) divides a received cryptocurrency payment into defined percentage parts and immediately transacts each part to nominated recipients — including tax agencies — without human intervention, with each transaction recorded in a blockchain.
Each NDEV token carries additional attached codes:
NDEV = Fragment X + A (e-wallet) + B (recipient validity) + C (logistics) + D (taxation)
Fragment X is detachable and reattachable, enabling the token to re-enter circulation after a transaction without losing its NIW provenance.
2.5 World Digital Currency (WDC) — The Global Settlement Layer
The WDC is constructed as a neutral, algorithmic basket of all national NDEV token supplies:
1 WDC(t) = Σ_j w_j(t) · 1 NDEV_j(t)
The critical constraint governing the entire global ledger is zero-sum issuance:
Σ_j ΔS_j(t) = 0
No country can gain by printing. No reserve-currency asymmetry exists. Settlement is automatically balanced.
3. Mathematical Proof: War Becomes Economically Irrational
Under the NDEV/WDC architecture the variables driving war profitability are structurally eliminated:
E = 0, D = 0, S = 0, M → 0, P → 0
Therefore:
B_w = f(0, 0, 0→, 0, 0→) ≈ 0
Since C_w (the real cost of war — human, material, infrastructural) remains strongly positive:
B_w - C_w < 0
The war probability function collapses:
p_war = σ(B_w - C_w) → 0
Core Engineering Statement: If economic value is created from real resources rather than issuance, B_w → 0 and p_war → 0.
4. Skydle Long-Range Probability Table
Structural probability estimates for key global risk categories. Values represent estimated structural probability (0–1).
| Code | Risk Category | 3yr | 5yr | 10yr | 20yr |
|---|---|---|---|---|---|
| P1 | Major currency conflict (USD-CNY-BRICS) | 0.68 | 0.74 | 0.82 | 0.90 |
| P2 | Large-scale military escalation between major blocs | 0.38 | 0.44 | 0.55 | 0.65 |
| P3 | Global debt crisis / disorderly reset | 0.57 | 0.63 | 0.70 | 0.80 |
| P4 | New neutral settlement experiments (NDEV-type pilots) | 0.05 | 0.10 | 0.21 | 0.40 |
| P5 | NDEV/WDC adoption in at least one regional corridor | 0.02 | 0.05 | 0.14 | 0.35 |
| P6 | If adopted: probability conflict incentives reduce | 0.85 | 0.85 | 0.85 | 0.85 |
| P7 | Business-as-usual leads to further escalation | 0.65 | 0.71 | 0.78 | 0.85 |
P6 (0.85) reflects the structural elimination of war's economic profit function as demonstrated in Section 3. This value is not optimistic; it is architectural.
5. Patent Foundation and Prior Art
Patent 1: AU2019101249 (Filed 14 October 2019)
Title: A Legal Tender to replace Fiat Currencies with National Digital Equivalents of Value (NDEV) during financial transactions.
Key claims: (1) Tokenisation of NIW into NDEV using geolocation parameters and gold-weight proportional value expression; (2) CST-enabled splitting of NDEV Fragment X with detachable codes for taxation and logistics; (3) NDEV usable in parallel with fiat currency as a transitional legal tender.
Patent 2: AU2018100999 (Filed 17 July 2018)
Title: Integration of Cryptocurrency Splitting Technology, Highly Scalable Blockchain Network and Mirror Advertising Technology for transparency of Business Operations and Taxation.
Key claims: (1) CST software divides cryptocurrency payments into percentage parts and instantly transacts to all recipients including taxation agencies without human intervention; (2) HSBN provides a scalable, hybrid central-decentralised blockchain with dynamic layer creation; (3) Mirror Advertising Technology (MAT) with GUID-based referencing enables searchable, auditable financial records across distributed databases.
Both patents are registered with IP Australia. Priority dates establish authorship of core mechanisms from 2018–2019, predating major CBDC developments globally.
6. Transition Scenarios: Plans A, B, and C
Plan A — Diplomatic: Voluntary adoption by major reserve-currency powers through multilateral agreement. Probability: low in short term. Requires political willingness to abandon issuance privilege.
Plan B — Crisis-Driven: Adoption forced by severe global debt crisis or disorderly reserve-currency fragmentation. NDEV/WDC becomes a survival tool for cross-border settlement. Probability: medium-high within 10 years given P3 trajectory.
Plan C — Survival Protocol: Deployment in conditions of near-total banking and communications breakdown. Local NDEV minting from real resource ledgers. Community clearing cells. Designed to prevent civilisational regression to barter or feudalism.
The three plans collectively demonstrate that the NDEV/WDC architecture is not a single-condition proposal. It is a resilient, scalable framework that becomes relevant across the full spectrum of global financial stress — from voluntary reform to civilisational emergency.
7. Role of Artificial Intelligence in Implementation
The NDEV/WDC architecture is inseparable from AI capability in four specific ways:
Resource Valuation: AI systems can calculate and update NIW valuations in real time across territories, replacing human-managed pricing with transparent algorithmic assessment.
Basket Weight Optimisation: The WDC basket weights w_j(t) require continuous rebalancing based on verified resource data. AI provides the computational infrastructure for this function.
HSBN Consensus Management: The dynamic layer-creation logic of HSBN — which scales blockchain capacity based on network parameters — is an AI-assisted adaptive system.
Conflict Probability Monitoring: The Skydle Probability Table is designed as a live instrument, updated by AI systems in response to macro conditions, providing decision makers with a continuously recalibrated risk dashboard.
For these reasons, the NDEV/WDC framework is submitted specifically to AI safety and monetary policy research bodies. The architecture requires AI to function at scale — and AI systems developing monetary reasoning capabilities require an architecture of this kind to avoid replicating the conflict-generating incentive structures of existing fiat systems.
8. Conclusion and Invitation to Review
The core proof:
If B_w = f(E=0, D=0, M→0, S=0, P→0) then B_w - C_w < 0 and p_war → 0.
This is not a claim about human nature or political will. It is a claim about architecture. If the monetary architecture removes the profit from war, war ceases to be a rational strategy for any actor within that system.
The authors invite independent technical review of:
- The mathematical formalism (Mathematical Appendix A)
- The patent specifications (AU2019101249, AU2018100999)
- The engineering commentary (v5 Engineering Commentary Appendix, 11 December 2025)
- The probability model (Skydle Long-Range Probability Table, 2025 Edition)
All materials are available at www.ramsmile.com.
References
file: MD5_NDEV_WDC_Technical_Brief.md
Date of Issue: May 2026 | UplitAU Pty Ltd, Sydney, Australia | www.ramsmile.com
Patent priority dates on record with IP Australia: AU2018100999 (17 July 2018), AU2019101249 (14 October 2019).
This document is issued as an open technical submission for independent review. It does not constitute legal advice, financial advice, or political advocacy.
© 2019–2026 UplitAU Pty Ltd / Skydle. All rights reserved.

