What the Stewardship Charter does, how each of its mechanisms works, why it binds its own author before anyone else, and what sits beside it that no clause can write.
Let me start with the sentence every critic thinks they have caught me on.
That is not an accusation. It is a quotation. I put it there myself. And a few lines later:
Large infrastructure usually answers the risk of being taken with insurance, investment treaties and friends in the ministry. When a new government decides to expropriate anyway, those protections are often worth very little.
The Charter takes a different route. It states, in public, that it cannot stop a sovereign government from doing what sovereign governments can do. Then it sets out a structure in which doing it costs the government the thing that made the asset valuable.
To see how, you have to see what kind of document this is.
Ask two questions of any document that governs your life: who wrote it, and who does it bind?
A loan agreement is drafted by the lender and binds the borrower. A lease is drafted by the landlord and binds the tenant. The terms of service you accepted this morning were drafted by a company's lawyers and bind you. As a rule, the stronger party writes, and the weaker party is bound.
The famous exceptions are documents that bound the powerful, and they were imposed from outside. In 1215 the barons of England forced King John to accept Magna Carta, and clause 61 appointed twenty-five barons with the right to seize the king's castles, lands and possessions if he broke it.1
The Stewardship Charter was written by the founder of the system it governs, and its constraints apply to that founder as fully as to anyone.
Guarantee by structure rather than promise by personality. Eight words, and the rest of the document follows from them.
I wrote those words knowing they apply to me. Part Fifteen says no single person may unilaterally alter pricing constraints, ceilings, water schedules, status rules, verification requirements or labelling standards. That includes the person who wrote it. A system that works only while I control it would solve nothing. So the Charter does not rely on me.
The Charter complies with local law. Participation is voluntary: anyone can decline. Where local law or host charges make a Charter requirement impossible, it does not fight; it withholds the CTMP Verified label from the affected delivery chain and records the limitation publicly.
Its leverage rests on something no sovereign controls: the right to say which outputs may carry the CTMP name and the CTMP Verified label, and on what terms. A government can seize a turbine. It cannot make an independent attestor certify a retail price that breaks the ceiling. It can tax. It cannot make the effect of its taxes on the household price disappear from the public record.
The Charter rests on four platform invariants. Each one closes a door that most systems leave open.
Invariant Two carries its own explanation:
Most governance manages the later links: regulating influence, investigating capture, capping rent. Invariant Two removes the first link platform-wide, including in every vertical. Ordinary unsecured trade payables that grant no control rights are excluded from the prohibition.
Invariant Four ends with a sentence most businesses could not live with:
Much of how large deals work is the negotiation of exceptions: the better rate for the biggest customer, the side letter for the best-connected investor. On pricing, ceilings, audit and financing, the Charter gives one answer to every such request. And Part Fifteen removes the people who could change that answer: changes that would weaken an invariant, raise a ceiling, reduce auditability or introduce discretionary pricing power are not permitted regardless of approval.
Most extraction hides inside structure. A service fee. A connection charge. A recovery surcharge. An affiliate billing company. A bundle. Each can be defended on its own, and together they turn a cheap input into an expensive bill.
The Charter does not argue with any of them. It looks at one number: the total the household pays, every charge from every source included. Core electricity leaves the metered delivery point at a posted US$0.025 per kilowatt-hour. The public default ceiling, for power sold as CTMP Verified, is US$0.05, all in.
Illustrative chains in US cents per kWh. The Charter does not ask how the price was built. It asks whether the total the end user pays stays within the ceiling.
Chain A can carry the CTMP Verified label. Chain B cannot, and the reporting record says why. Nobody is fined and nothing is switched off. Chain B loses the label, and with it eligibility for expansion and promotional attestation, until its retail schedule is back within the ceiling.
Governments keep every right to tax. If taxes alone push a chain over the ceiling, the country stays in the network; that chain simply cannot be called CTMP Verified.
Cap the price to the customer, and the margin tends to move between divisions, where fewer people look. The steel division overcharges the turbine division. The water division bills the housing division a handling fee. The outside price holds, and the rent comes back inside.
Part Nine closes that door. Transfers between CTMP verticals follow published rules designed to pass costs through without markup, and an internal toll that behaves like rent is treated as a Charter violation.
It is important to be exact about the boundary. Part Nine governs transfers between verticals. Part Ten governs sales of non-essential vertical outputs to the outside world, and it does not cap them: this Charter does not prohibit high margins on non-essential vertical outputs. It makes them visible. CTMP's own external sales follow a published pricing corridor designed to stay below incumbent medians.
My broader platform policy, which sells food and everyday goods inside the platform at zero margin and which drives the lettuce in my cost model from 98 to 53, is a design rule I have set on top of the Charter. It is not a Charter requirement, and I keep the two separate.
Most enforcement ends, eventually, in force: the fine is backed by the court, the court by the bailiff. In utilities the equivalent is the switch. Pay, or we cut you off.
The Charter enforces through two systems instead: network status, which decides eligibility for CTMP terms, integrations and expansion, and verification status, which decides which outputs may be called CTMP Verified.
Compliance carries full CTMP terms. Continuity keeps essential-use and life-safety delivery flowing but suspends membership benefits. Exit follows only seizure, override or nationalisation, or persistent misrepresentation combined with refusal of auditability. Re-entry requires restored compliance, independent verification and a public disclosure. Status rests on documented facts, not negotiation.
Beside network status sits the label. An output is CTMP Verified, Not CTMP Verified, Disputed, or Out of Network, each a public statement of fact about a specific delivery chain. The Out of Network label is a statement of fact, not a judgment.
Rules are only as good as whoever applies them, and in most institutions that is a person who can be persuaded, pressured or paid.
The Charter assigns enforcement to the Sovereign Logic Engine, a deterministic system that applies the constraints automatically. The word sovereign is defined narrowly: non-discretionary enforcement and immunity to capture, not any claim of state authority.
There is one sealed emergency safeguard, for catastrophic malfunction, verified code corruption or critical compromise. It can only place the system into a defined safe mode. It cannot modify a tariff, a ceiling or a rule, and any activation is logged, attested and disclosed.
Stewardship Charter, Part Fourteen: the four layers work together. Pricing compliance is checked through published retail schedules, attestation and audit sampling, with no surveillance of individual households.
In most systems, ambiguity is where money hides. Here, ambiguity counts against whoever created it.
In the Odyssey, Ulysses knew that when his ship passed the Sirens he would want to steer onto the rocks. So he had his crew bind him to the mast before they got there.
Part Fifteen does something similar, with a precise boundary.
The rest of the document can change, but only one way. Non-invariant provisions may be amended with documented justification, multi-party approval, public notice and tamper-evident logging of the change itself. The public edition says it will be refined for clarity, definitions and implementation detail, with every change dated and logged. What no process can do is weaken an invariant, raise a ceiling, reduce auditability, introduce discretionary pricing power or compromise the labelling system.
So the mast is not the whole ship. The sails can be trimmed in public. The course cannot be changed toward the rocks.
Imagine a government, some years after a facility is running, decides it is too valuable to leave under anyone else's terms. It takes control.
The Charter's response is immediate and recorded: Exit. The facility no longer receives CTMP terms, integrations, supply chain access, technical support, attestations or ecosystem benefits. Its output is labelled Out of Network and may not be represented as CTMP priced, CTMP Verified or Charter-governed.
Nothing is switched off. Exit status does not prevent physical operation.
In the platform's design, a facility is one node in an integrated system. Its replacement parts, turbines, tunnelling equipment, materials, software support and trained people come through the network. After Exit, the operator has to source all of that elsewhere, on the market's terms.
The workforce has a choice too. Under Part Twelve, CTMP offers affected personnel the option to transfer to other network facilities, subject to lawful work authorisation, host-country requirements and their own choice. Those who stay may stay; the Charter does not prevent it.
Vertical operations in the network at 17 per module: power, water and 15 integrated verticals, per the author's programme design target of 80 to 100 modules. A facility in Exit loses CTMP terms, integration and supply chain access across all of them.
Exit is not permanent. There is no permanent exclusion for seizure or for past misrepresentation. A participant may apply to return by restoring CTMP terms, metering, published schedules and independent attestation, with a public disclosure of what happened and what changed.
A fair reader will object. I have spent years criticising gatekeepers, who make others dependent and then charge a toll. Is a network that becomes more costly to leave not a gatekeeper too?
| A gatekeeper | The Charter | |
|---|---|---|
| Uses dependence to | Change the terms: raise the rent, reprice, tighten covenants | Cannot. The posted tariff is fixed; ceilings cannot be raised by any process |
| Offers the powerful | Better deals, side letters, priority | The same published terms as everyone, by invariant |
| Enforces by | Cutting you off | Status and labels; life-critical supply is never used as a weapon |
| If you leave | Often punishes permanently | No permanent exclusion; re-entry on restored compliance |
| Hard to leave because | Staying is expensive and leaving is punished | Leaving means giving up terms that stay the same for everyone |
A gatekeeper uses dependence to change the deal. The Charter removes anyone's power to change the deal. What makes leaving costly is only what the leaver gives up.
I hold myself to stating where the argument is weakest.
The cost of exit grows with the network. At the start there is barely a network. The sequence is fixed: a country that reaches its required count on The Wall earns a formal review, not a power plant. If the review and the host government both say yes, the first build is a 50 megawatt proof of concept. Its measured results carry the programme to 250 megawatts, and from there module by proven module.
In those early years there are no neighbouring modules, no large integrated supply chain, no deep bench of transfer options. The asset is growing, and the web that makes exit costly is thin.
Careful host agreements help. Designing early facilities to depend on the network helps. Neither fully answers the question.
Hold that thought.
A government can take the concrete. It cannot take the network.
It can tax. It cannot make the effect on the household price disappear from the record.
It can keep the machines running. It cannot keep the CTMP terms, the integration, the attestation or the people who choose to transfer.
The older way to protect an asset was guns, lawyers and switches.
The Charter protects it with terms that are identical for everyone, and a public record of what leaving would cost.
That is where I meant to end.
But it does not answer the question I told you to hold.
The Charter has eighteen parts. I wrote all of them. Part Eighteen is a canonical public statement, the paragraph anyone may use to describe the document accurately. The text is written, and any future change to it is bounded by Part Fifteen.
One part is different in kind. Part Seventeen describes The Wall. Every other part tells the system what to do. Part Seventeen describes something the system does not produce and cannot fill in: a public ledger of demand.
The Charter is careful here, and so am I. The Wall creates no legal relationship. It is not a petition to any government. Notifications to people who opt in are informational and are not a call to action. Donations buy nothing: the relationship between donation and CTMP access is zero. The Wall does not enforce anything.
So what follows is my argument, not a clause.
Go back to the weakness. Early on, the cost of exit is small. What exists from the very start, before the first shovel, is the public record: how many voices, recorded under that country's name, asked for the proposition to be tested, in the open, on a count anyone on earth can see.
Conceptual, not measured. The cost of exit, which the Charter creates, grows with the network. The public record of demand, which the Charter only describes, exists before the first facility is built.
A government that later took the facility would be taking something that a large, public, recorded demand asked to have tested. The Wall is anonymous, so no one can say who those voices were. But the number is public, and a government has to answer for it in the open. That is not a legal barrier. It is public standing, and in the years when the cost of exit is weakest, it may be the most important standing the project has.
The Charter is finished. The record beside it is not.
The great charters of history were backed by a small number of named, powerful people.2 This one is written the other way round. Its text binds the powerful, including its author. What stands beside it is an open count that anyone can add to, with no account, no name, no email and no money.
I wrote the eighteen parts.
The record is written by you.
The record has only ever grown one way: one person handing it to another. So here is the arithmetic, with the assumption stated.
Suppose everyone already on The Wall for a country brings in three new people from that country, once, and each of those three does the same, once. After one round the count is four times today's. After two, thirteen times. After three, forty. After four, 121. After five, 364.
Each voice recruits three new voices in the same country, once. Cumulative multiple after k rounds is (3 to the power k+1, minus 1) divided by 2. Required multiples from the Race Pulse snapshot of 28 to 29 September 2026: Papua New Guinea 188x, Ireland 209x, India 215x, South Korea 219x, Mexico 228x, Pakistan 237x, Laos 248x. Illustrative, not a forecast.
Every one of the eight leading countries needs between 188 and 248 times its current count to reach its required count. Four rounds falls short. Five rounds clears all of them.
For one person, five rounds of three is 1 + 3 + 9 + 27 + 81 + 243: 364 voices that trace back to a single decision.
Reaching the count earns a formal review, not a facility. But nothing moves until the count does.
The barons held a king to his word with twenty-five names.
The Charter holds its own author to his word with eighteen parts.
What it cannot do is create the demand that gives it standing. That part was never mine to write.
Add the next line.
Make your country count. No account. No name. No email. No money.
The Wall is anonymous and binds no state; reaching a required count earns a formal review, not a deployment. Wall total from the public count on 29 September 2026. Nothing described in the CTMP design has been built.