The economy
Economic Constitution
The thirty-four laws the KRED economy runs on. Everything else is subordinate.
Thirty-four foundational laws: I to XX original, XXI to XXV added by Amendment A01, XXVI to XXXIV added by Amendment A02. Everything else in this repository is subordinate to them. A default value may be tuned; a law may not be tuned. It can only be amended deliberately, with a version bump and a stated reason.
Each law below is stated, then explained: what it means in practice and what it prevents. The "prevents" column is the important one, every law here exists because something breaks without it.
Law I: Official Issuance
Official KREDS can only be issued by the Kreds Network.
Means: every official KRED originates from exactly one of three sources: the Central Bank reserve, another existing holder, or an explicit reversal/refund of an earlier transaction. No KRED appears from nowhere.
Prevents: local minting. A self-hosted instance, an org admin, or a compromised client cannot create official supply. The maximum supply is only meaningful if issuance has a single choke point.
Law II: Auditable Movement
Every KRED movement must have an auditable ledger entry.
Means: balances are derived, never stored-and-mutated. user.balance += 30
is a bug, not an optimisation.
Prevents: unexplainable state. If a balance cannot be reconstructed from its entries, fraud investigation, reversal, and supply conservation all become impossible at once.
Law III: Productive Creation
Shipping may create value. Reviewing primarily circulates value.
Means: merges mint (bounded, quality-scored). Reviews transfer from author to reviewer. The economy grows through shipped work and moves through helping others ship.
Prevents: infinite inflation via review farming. If reviews minted, two accounts reviewing each other would print money forever.
Law IV: Organization Boundary
GitHub-derived economic activity first belongs to the economy of the connected GitHub Organization.
Means: value earned from work inside an org lands in an org-scoped position before it is anything else even when the org uses official KRED 1:1.
Prevents: earn-here-spend-there escapes. The boundary is what makes reversals, debt, collusion detection, and settlement enforceable.
Law V: Same Currency Does Not Mean Same Accounting Context
A team using KRED 1:1 still requires organization-scoped positions, settlement, debt, and risk controls before KRED becomes globally withdrawable.
Means: "we just use KRED" is a currency choice, not an accounting choice. The org position exists regardless.
Prevents: the obvious shortcut, treating 1:1 teams as a direct pipe into the global wallet, which would delete every protection in Part XI.
Law VI: Debt Is Allowed
Users may hold negative net positions within defined credit limits. Actual KRED balances are never negative.
Means: owing more review than you have paid for is a legitimate game state, not an error condition. The debt is recorded as a liability alongside a KRED balance that stays at or above zero. The intended way out is productive work, especially reviewing others.
Prevents: a dead economy. If authors could not owe, review requests would stall the moment a balance hit zero, and the core loop would deadlock.
Amended by A01 (
v0.2). Previously read "Users may hold negative positions", which permitted negative currency balances, and therefore let two accounts at zero mint spendable KRED by reviewing each other. See Law XXI and 23 Review Funding, Debt and Credit.
Law VII: Extraction Is Not Guaranteed
Debt, pending value, borrowed value, or unsettled activity cannot be exported as global settled KRED.
Means: withdrawable is a strict subset of balance. Value must survive the settlement window and risk checks before it leaves the org context.
Prevents: the core farming attack, mint locally, export instantly, default locally, walk away.
Law VIII: Debt Is Repaid First
When an economic position carries outstanding debt or unfunded receivables, future eligible earnings settle those obligations before becoming withdrawable or transferable.
Means: a user owing 200 K who earns +40 K now owes 160 K and holds
nothing new. That +40 K cannot be routed anywhere else. Where the obligation is
an unfunded receivable, the reviewer is paid before the author receives a
kredbit.
Prevents: parallel-account debt parking: earn on the indebted account, route the earnings out, leave the hole permanently open.
Amended by A01 (
v0.2) to cover receivables and non-organization debt, and to state the ordering explicitly:earnings → debt → pending settlement → available → withdrawable.
Law IX: Global KRED Belongs to the Holder
Once settled, Official KRED belongs to the user's global wallet and follows them across the network.
Means: one human, one global KRED wallet, regardless of how many orgs they belong to. Leaving a team does not confiscate settled KRED.
Prevents: employer-captured reputation. Kreds is a portable developer economy or it is a corporate points program; it cannot be both.
Law X: Local Currency Stays Local
Organization-specific currencies belong to their respective economies.
Means: ZIT is ZitDevs' currency. It does not travel, and holding 4,000 ZIT
never means holding KRED.
Prevents: implicit convertibility. Without this, every org currency becomes a synthetic claim on the global reserve.
Law XI: Independent Economies Are Free
Independent self-hosted economies may define their own currency and policy, but cannot mint or impersonate Official KRED.
Means: run your own instance, print a billion of your own currency, set your own fees. That is your economy. It is simply not the Kreds Network.
Prevents: counterfeit legitimacy, a self-hosted instance advertising local tokens as network-backed.
Law XII: Network Membership Is Optional for Self-Hosting
Self-hosting does not require Kreds Network participation. Joining later is permitted through a defined migration process.
Means: the open-source path is genuinely usable standalone; the network is an opt-in upgrade with registration, signed events, and identity verification.
Prevents: a fake open-source story where the software is useless without the hosted service.
Law XIII: Joining Does Not Rewrite History
Joining Kreds Network gives an existing local economy a reserve relationship; it does not erase its previous balances or ledger.
Means: a member's 42,000 local units before joining are still 42,000 after. The join is recorded as a snapshot: opening supply, opening reserve, opening backing ratio.
Prevents: migration as a stealth wealth reset, the fastest way to lose the trust of every existing member.
Law XIV: Reserve Backing Is Not Fiat Value
KRED-backed local currencies may have relative backing against KRED without Kreds assigning KRED a cash price.
Means: publish 1 ZIT = 0.025 KRED. Never publish 1 KRED = $0.12.
Prevents: turning a gamification layer into a financial instrument, with every legal and adversarial consequence that follows.
Law XV: Rules May Change, History May Not
Economic rules are versioned and forward-only.
Means: every transaction stores the rulesVersion that produced it. Raising
the review maximum affects future eligible activity only.
Prevents: retroactive re-pricing, and the silent policy change, the single fastest way to destroy trust in any economy.
Law XVI: Bots Are Not Developers
Bots, GitHub Apps, and AI agents do not receive human economic rewards.
Means: every GitHub identity is classified HUMAN | BOT | AI_AGENT | UNKNOWN.
Only eligible humans participate economically. Dependabot, Renovate, Copilot and
Claude Code may appear in history at 0 KRED.
Prevents: the most trivially automatable farm in the entire design.
Law XVII: Unclaimed Identity Can Have History
A GitHub identity may earn verified KRED before claiming a Kreds account.
Means: your Kreds history starts before your Kreds account does. Review someone's PR without ever signing up, and the value is waiting for you.
Prevents: the cold-start problem, and the unfairness of rewarding signup date over actual contribution.
Law XVIII: Unclaimed Accounts Are Passive
Unclaimed identities cannot perform voluntary economic actions.
Means: an unclaimed identity can receive verified GitHub-derived value. It cannot send, donate, exchange, withdraw, or create economies. Voluntary transfers to unclaimed identities are also blocked.
Prevents: phantom-account farming, spraying value into fabricated identities and claiming them later.
Law XIX: Every Incentive Is Adversarial
Every reward mechanism must be designed under the assumption that someone will eventually attempt to farm it.
Means: the design question is never "is this a nice reward?" It is "what does the cheapest possible abuse of this reward look like, and what stops it?"
Prevents: shipping incentives whose failure mode is only discovered in production, by the people exploiting them.
Law XX: Economic Integrity Over Convenience
No UX shortcut may bypass ledger integrity, settlement rules, fraud protections, or supply conservation.
Means: "let's just credit it instantly so the UI feels snappy" is not a trade-off to be weighed. It is a violation.
Prevents: integrity erosion by a thousand small product decisions, each individually reasonable.
Amendment A01: Review funding and debt
Added by policy
v0.2. These five laws close the hole that negative currency balances opened: a review paid out of an overdraft was, in effect, minting. See 23 Review Funding, Debt and Credit.
Law XXI: No Monetary Creation Through Debt
A negative economic position may represent a liability, but it may never create spendable Official KRED. Every settled reviewer reward must be funded by existing Official KRED.
Means: balance >= 0 at all times, for every account. Debt is tracked
alongside the balance, never inside it.
Prevents: the founding accounting bug. Under the old model, two accounts at zero could review each other and produce spendable KRED that never came from the 5,000,000 supply, indistinguishable, once it existed, from legitimately issued currency. The cap was not a cap.
Law XXII: Reviews Must Be Funded
A valid Code Review may create an economic claim, but it becomes spendable KRED only when funded by the author's settled balance, a Review Fund (including Treasury KRED explicitly allocated into one), the Kreds Review Credit Facility, or a bounded Central Bank program (platform-funded review rewards and unclaimed-debt protection payments), subject to the same eligibility gates. New funding sources may be added only by constitutional amendment.
Means: the reviewer always earns. What varies is whether they are paid now or hold a claim until funding arrives. Value is never invented to close the gap and the list of places value may come from is closed. A Treasury funds reviews only through an explicit, ledgered allocation into a Review Fund; it is never an automatic step of the waterfall.
Prevents: the two bad alternatives, inventing KRED to pay the reviewer, or telling the reviewer their work was worth nothing because the author was broke. The closed enumeration also prevents a quieter failure: an "other legitimate mechanism" clause in an adversarial economy (Law XIX) is a door someone eventually walks through.
Amended by A03 (
v0.4). Previously listed "a Treasury" as a direct source, which contradicted the waterfall's requirement of an explicit Review Fund allocation, and ended with the open clause "or another legitimate funded mechanism".Corrected in A03 audit round 2. The first closed enumeration accidentally excluded the two Central Bank programs that legitimately pay reviewers: platform-funded rewards (reviewer #4+, simple re-approval) and protection payments, thereby forbidding flows that Law XXIII's gating and chapter 25 explicitly regulate. The list now names them; it remains closed.
Law XXIII: Central Bank Credit Uses Existing Reserves
The Kreds Review Credit Facility may temporarily finance eligible review activity using existing Central Bank reserves. Credit does not increase the official KRED supply. Credit draws require repository economic eligibility and an eligible reviewer.
Means: financed reviews move KRED from reserve into circulation and record a matching debt. Total supply is untouched; circulating supply rises, which is precisely why the facility must be capped, and why a draw is gated exactly like a minting path: from the reserve's perspective it is the same act.
Prevents: an unbounded subsidy, and reserve extraction through collusion, a throwaway repository full of fresh accounts drawing credit to pay one real reviewer, then abandoning the debt. Without the eligibility gate, the facility would be the cheapest farm in the system.
Amended by A03 (
v0.4) to add the eligibility clause. The A02 principle (every path that moves reserve KRED into circulation gets the same gate) covered platform-funded rewards but not credit draws.
Law XXIV: Unfunded Work Is a Claim, Not Currency
An economically valid but unfunded review creates a receivable. Receivables do not count toward KRED supply and cannot be transferred, spent, or withdrawn until funded.
Means: a receivable is recorded, visible, and settled ahead of the author's own future earnings, but it is not money and never appears in circulating supply.
Prevents: a shadow currency. A transferable claim would be a second money supply with none of the first one's controls.
Law XXV: Reciprocity Is Economic, Not Behavioral
Kreds does not require every developer to review as much code as they submit. Teams and projects may specialize, but persistent review deficits must be financed by productive activity, treasuries, sponsors, or limited credit.
Means: a maintainer who ships constantly and reviews rarely is doing a real job, not gaming anything. The obligation to balance sits with the project, not with each individual.
Prevents: enforced tit-for-tat, which would destroy role specialization and punish exactly the people carrying the most project responsibility.
Amendment A02: Contribution Points and economic eligibility
Added by policy
v0.3. These nine laws split recognition from issuance and close the last path by which one person acting alone could mint Official KRED. See 24 Contribution Points and 25 Repository Economic Eligibility.
Law XXVI: Contribution Is Not Currency
Contribution Points represent verified work and reputation. They cannot be transferred, spent, exchanged, or used as KRED, and they have no fixed or implied conversion rate into KRED in either direction.
Means: points have no supply, no ledger position, and no economic effect. They never appear in the supply equation, and no exchange, redemption, or "points buy" mechanic may exist.
Prevents: a second money supply with none of the first one's controls, no cap, no ledger discipline, no settlement, no funding requirement. A conversion rate would make every supply control bypassable by minting reputation.
Law XXVII: Contribution Does Not Decrease
Contribution Points are cumulative historical recognition and do not decrease through spending, debt, or normal economic activity.
Means: paying for review does not erase the record that you did the work. Points remain adjustable when the underlying contribution is invalidated a revert, confirmed fraud, or an actor reclassified as a bot.
Prevents: a reputation score that punishes participation in the economy it is attached to.
Law XXVIII: Monetary Eligibility Requires Stronger Verification
An activity may earn Contribution Points without being eligible to create or transfer Official KRED. KRED issuance requires a higher standard of economic verification than Contribution Point recognition.
Means: recognition and issuance are separate decisions with separate evidence, and the bar for money is deliberately higher than the bar for credit. When the two standards conflict, the economy defers and reputation proceeds.
Prevents: the forced choice between insulting legitimate contributors and monetizing unverifiable work and, more importantly, the supply being governed by the weaker of two evidentiary standards.
Law XXIX: Self-Directed Private Merges Do Not Create KRED
A merge performed in a private repository without a valid eligible human review does not create Official KRED.
Means: owning the repository, opening the PR, and merging it yourself is not evidence of anything Kreds can verify.
Prevents: the last remaining unilateral minting path, one person, one private repository, unlimited repetition.
Law XXX: Public Visibility Alone Is Not Sufficient
Making a repository public does not automatically grant full monetary eligibility. Repository relevance and trust must be established.
Means: a public repository created ten seconds ago carries no more evidence than a private one.
Prevents: trivially converting the private-merge farm into a public-merge farm by flipping a visibility toggle.
Law XXXI: Repository Relevance Is Multi-Signal
GitHub stars may influence repository trust, but no single popularity metric defines economic legitimacy.
Means: trust is a score over many signals, moving gradually, never a threshold on one number.
Prevents: buying eligibility. Any single metric that unlocks issuance becomes a market.
Law XXXII: Human Review Can Establish Economic Validation
A meaningful Code Review from an eligible independent human may establish economic eligibility for work that otherwise lacks sufficient repository trust.
Means: a second party with something to lose is itself a form of evidence. This is the escape hatch that keeps eligibility usable for new and private work.
Prevents: locking legitimate teams out of the economy for the crime of working in a private repository.
Law XXXIII: Established Public Projects May Earn Without Mandatory Review
A sufficiently trusted public repository may qualify merged work for KRED issuance even when a specific Pull Request receives no formal Code Review.
Means: the repository's external history is the validation. Solo maintainers of widely used projects participate normally.
Prevents: punishing exactly the open-source maintainers the economy most wants to reward, for the structural fact that they have nobody to approve their PRs.
Law XXXIV: Alternate Accounts Cannot Legitimize Self-Directed Work
A user may not create economic eligibility by reviewing their own work through controlled alternate identities.
Means: the reviewer used to unlock eligibility must be a genuinely independent, trusted, human identity.
Prevents: defeating Law XXIX with a second GitHub account, which is otherwise the obvious next move.
Constitutional summary
| Law | One line |
|---|---|
| I | Only the Network issues official KRED |
| II | Every movement is a ledger entry |
| III | Ship to create, review to circulate |
| IV | Org activity belongs to the org economy first |
| V | Same currency ≠ same accounting context |
| VI | Debt is legal |
| VII | Extraction is not guaranteed |
| VIII | Debt is repaid first |
| IX | Settled KRED belongs to the holder |
| X | Local currency stays local |
| XI | Independent economies are free but not official |
| XII | Self-hosting needs no network |
| XIII | Joining preserves history |
| XIV | Backing is not a fiat price |
| XV | Rules change forward-only |
| XVI | Bots are not developers |
| XVII | Unclaimed identities can earn |
| XVIII | Unclaimed identities cannot spend |
| XIX | Every incentive is adversarial |
| XX | Integrity beats convenience |
| XXI | Debt never creates spendable KRED |
| XXII | Reviews must be funded to become currency |
| XXIII | Credit deploys reserves, it does not issue |
| XXIV | Unfunded work is a claim, not money |
| XXV | Reciprocity is economic, not behavioural |
| XXVI | Contribution Points are not currency, and never convert |
| XXVII | Contribution does not decrease from economic activity |
| XXVIII | Money requires stronger proof than credit |
| XXIX | Self-directed private merges do not mint |
| XXX | Public visibility alone is not enough |
| XXXI | Relevance is multi-signal, never one metric |
| XXXII | Human review can establish eligibility |
| XXXIII | Trusted public projects earn without mandatory review |
| XXXIV | Alternate accounts cannot self-validate |
This page renders content/economy/constitution.md from the repository. Same file, whether you read it here or after cloning. Suggest an edit.