Caper

Caper is a platform on Radix for creating and running member-owned organizations, which it calls capers. Each caper is created with its own token, a market that trades that token, and a treasury its holders govern, and a member can leave at any time with a share of what the treasury holds. It covers the whole life of a Decentralized Autonomous Organization (DAO): creation, fundraising, governance, exit and acquisition. It runs on Stokenet, the Radix test network, so its tokens and the XRD in its treasuries have no value; caper.network has said so on every page since 8 October 2026.

Summary

Caper sets out the case for its design in an essay, The Exit Right. In most DAOs, it argues, a member's voting power and their claim on the treasury are separate quantities, so votes can be gathered without committing capital and a claim can be held by someone who never governs. Caper ties the two together, and the essay draws on Albert Hirschman’s account of exit and voice and on the cooperative tradition studied by Elinor Ostrom.

Each DAO created on the Caper platform operates through a bonding curve mechanism that serves as an automated market maker, providing immediate liquidity for governance tokens and establishing deterministic pricing without traditional market-making intermediaries. The platform employs a governance system that calculates voting power from both token holdings and voting history, and settles a member's claim on the treasury against a non-transferable stake token that only voting mints. Since 29 August 2026 a proposal has had to survive two independent tests rather than one: a ranked ballot, and then a window in which the caper's own market can object.

Background and Development

Historical models

Caper's essay The Exit Right sets the design against three older answers to what a member keeps on leaving: a medieval guild, which gave almost nothing back; a worker cooperative of the kind Elinor Ostrom studied, which settles the member's stake; and a joint-stock company, whose shares can be sold at any time without the firm's consent. It argues that leaving a caper is severance in the joint-stock sense, and that the new part is tying it to governance, so that a member's vote never counts for more than the share they could take out.

Platform Features

Bonding Curve Mechanism

Caper's core technological innovation centers on its implementation of bonding curves as automated market makers (AMMs) for DAO governance tokens. Each caper operates through a mathematical function that determines token pricing based on circulating supply. This mechanism provides several advantages over traditional liquidity pools: permissionless operation, deterministic pricing, and liquidity at every point on the curve. It is not free of price impact. A purchase pays the area under the curve between the old supply and the new, so a large order pays a higher average price than a small one, and Caper's own page on the curve calls the common "zero slippage" description imprecise for that reason. What the curve fixes is the size of the impact: it depends only on the trade and the curve, and can be known before the trade is signed.

The bonding curve system eliminates common risks associated with externally supplied liquidity, including rug pulls and price manipulation that have plagued other decentralized finance protocols. By holding collateral directly within the curve contract, the system removes dependence on external market makers or liquidity incentives. Because the price rises with the supply sold, an earlier buyer pays less per token than a later one.

Cashtag System

The platform employs a unique identification system called "cashtags" - alphanumeric identifiers that prevent spoofing and establish clear project identity. The registration fee is ten raised to the power of seven minus the number of characters in the cashtag, so a six-character cashtag costs 10 XRD and every character dropped multiplies the fee by ten: 100 XRD at five characters, 1,000 at four, 10,000 at three, 100,000 at two and 1,000,000 at one.

A cashtag runs one to six characters and is uppercase A to Z or digits 0 to 9, and the contract rejects a payment that is even slightly off the figure its length sets. Two cashtags are reserved: XRD, because a governance token whose symbol is locked to its cashtag would be indistinguishable from the network’s own currency, and CAPER, which only the protocol administrator can mint. Caper publishes the ladder, and the create_caper method asserts the payment against it.

The scarcity imposed by unique cashtag requirements, combined with the perpetual liquidity guarantee provided by bonding curves, creates economic incentives for valuable cashtags to be "recolonized" rather than abandoned if founding teams depart. This mechanism provides community members with potential paths back to profitability even in cases of project abandonment.

Fixed Token Supply Architecture

All capers implement a standardized token supply of 100 billion governance tokens, which enables direct price comparisons between projects and prevents attempts to manipulate apparent market capitalizations through supply adjustments. This fixed supply model provides transparency in ownership percentage calculations and eliminates the Cantillon Effect associated with unbacked token issuance.

Governance System

Vote Weight Calculation

Caper implements a sophisticated governance mechanism that calculates voting power and treasury exit rights. This mathematical approach addresses persistent challenges in DAO governance, including voter apathy and free-rider problems. The function creates natural threshold mechanisms where meaningful voting power and exit rights accrue only to members demonstrating both economic commitment through token ownership and active engagement through consistent participation in governance processes.

The vote weight system provides protection against governance attacks, as temporary token acquisition without corresponding voting history cannot immediately grant disproportionate influence. This prevents the type of governance sniping that has affected protocols like Rook DAO, where attackers acquire large token positions specifically to manipulate governance outcomes.

Ordinal Ranking Voting

The platform implements ordinal voting systems for all governance decisions, where voters rank options by preference rather than selecting single choices. In a ballot with multiple options, the highest-ranked choice receives points equal to (number of options - 1) multiplied by the voter's vote weight, with subsequent choices receiving proportionally fewer points. This method elicits more comprehensive information about voter preferences compared to simple majority systems while avoiding the complexity of pairwise comparison methods.

Supermajority Requirements

A ballot carries between two and five options, and one of them is always a "Do nothing" entry, so the smallest slate is a single answer standing against a rejection. The leading option passes only if it takes 1.5 divided by the number of options of the weight cast, and only if it is not the "Do nothing" entry, which is barred from winning even when it clears the bar.

That share is measured against the ranked points cast rather than against the voters: a full ranked ballot of x options awards x(x-1)/2 points, of which any one option can take at most x-1, so the most an option can score is 2/x and the threshold of 1.5/x is three-quarters of that ceiling at every length of slate. In figures, two options need 75% of the points cast, three need 50%, four need 37.5% and five need 30%. Caper states the threshold and the shape of the slate.

The Market Window

Clearing the ballot is not the whole test. Since 29 August 2026 the winning option has to survive a second phase in which the caper's own market can object. Someone triggers the winner, which locks the caper's trailing average token price as a baseline and opens a market window, and the treasury performs the action only if the average price across that window finishes at or above the baseline. Holding through the window counts as consent, buying as support, and selling is the objection.

Caper's own account of the mechanism dates this and the two arrangements before it: until 26 August 2026 a caper decided by ranked ballot alone, from 26 to 28 August the ballot was removed and price decided on its own, and the deployment of 7 September 2026 keeps both phases and measures the window as an exact price-time integral.

Proposals That Act and Proposals That Debate

Most options on a proposal carry an action the treasury performs if the option wins and survives the market window: a payment to any Radix account, an investment in another caper's token or the sale of one, a change to how the caper's own token is presented, and, for the platform's own caper only, an upgrade of the platform's logic. A debate option carries no action. It is ranked and can win like any other, and there the proposal ends, with no market window because there is nothing to execute. Caper's page on paying from a treasury lists each kind against the contract.

Creating a proposal and casting a ballot each cost a fee, and both are paid into the caper's own treasury rather than to the platform; the creation fee is kept whether the proposal passes or fails. There is no turnout quorum. An option passes on its share of the weight actually cast, so a proposal few members care about can pass on a small vote, and it still has to survive the market window.

DAO Lifecycle

Creation

Each DAO begins with an empty treasury that serves as the repository for value generated through organizational activities and operations. The treasury functions independently from the collateral backing the bonding curve, representing wealth accumulated through the DAO's productive efforts rather than token trading activities.

Every caper also gets a Trollbox, a chat room on its profile with a separate room under each of its proposals. Holding any amount of the caper's token, its vote token or its founder badge is enough to post; there is no minimum and no governance setting for one. The founder moderates, and since September 2026 members can collapse a line by flagging it, each flag weighted by the same vote weight that decides a ballot, so a holder who has never voted can post but cannot flag. Caper describes the room and how its flags are counted.

Trading and Market Development

As capers mature, their governance tokens become available for trading through both the primary bonding curve market and secondary peer-to-peer transactions. A swap moves a position from one caper to another in a single transaction: the contract sells the tokens on the first caper's curve and puts the XRD it pays straight into a buy on the second. It is two trades rather than a direct exchange, so each leg pays its own caper's skim, and Caper prices it that way.

Fundraising and Growth

No vesting schedule holds a founder’s allocation back, because there is no allocation to hold. The founder’s claim accrues trade by trade out of the slice taken on each purchase and sale, into two vaults the caper keeps for it, and the founder withdraws from those vaults through withdraw_founder, which checks that the caller presents that caper’s founder badge and that the amount asked for is no more than has already accrued. There is no milestone, no cliff and no price gate on the way out. Caper puts it the same way: what stops a founder emptying the reserve is not a vesting schedule but the absence of any method that would do it. The badge is a transferable bearer token, so the stream it collects can change hands without a proposal.

An existing token's community can move onto Caper through what Caper calls a vampire caper: a caper launched with the existing token's resource address as one argument, which can never be added or changed later. Any holder of that token can then hand it in at any time, without a fee, for caper tokens at a rate fixed at launch, 100 billion divided by the origin token's supply, and what they hand in is locked away for good. Only one caper can ever absorb a given token. Caper documents the binding and the rate.

No separate fund selects projects to back. What connects the platform to every caper launched on it is a slice of each trade, taken on buys and sells alike in the same contract call that settles the trade. Roughly half of it goes to the traded caper's own treasury in XRD; the founder takes about as much, mostly in XRD and partly in the caper's tokens; and a small token slice goes to the Commons, which is the treasury of the platform's own caper.

There is no separate trade fee: since a redeploy on 25 September 2026 the treasury's share of that slice replaces the flat fee a trade used to pay. The platform accrues its position in a caper trade by trade rather than being granted one, and the split is documented against the contract method that performs it.

Governance

Every action a treasury takes, from a payment to an investment, goes through the same ranked ballot and market window described above. There is no separate class of binding proposal to graduate to, and no delay set aside for dissenters, because a member who disagrees can exit at any time rather than inside a window. Each ballot a member casts mints a stake token, so the members who vote most carry more weight in later ballots and a larger claim on the treasury when they leave.

Treasury Development and Value Accumulation

A caper's treasury fills from its share of the slice taken on every trade of its token and from the fees members pay to open a proposal and to vote, which the contract deposits into it. It spends only through a proposal that passes: a payment to a recipient, an investment that buys another caper's tokens, or a divestment that sells them. Caper lists the proposal kinds.

Exit and Dissolution

A member leaves by calling the caper's exit method. It is not a proposal, so it is not voted on and needs nobody's approval. The contract checks that the member hands in some of that caper's governance tokens and some of its non-transferable stake token, and that the transaction satisfies the owner rule of the account the payout goes to, so only that account's owner can take it out.

Since a redeploy on 21 September 2026 it also checks that the stake tokens are the member's own: the same transaction first calls exit_open, which reads the account's stake-token balance, and exit aborts on Vote tokens must all come from the recipient unless the tokens handed in equal the fall in that balance.

Voting is the only thing that mints a stake token: one for each ranked ballot cast. A member who has only ever bought holds none, and the call aborts on Exit requires vote tokens rather than paying out nothing, so leaving requires having voted at least once. Trading minted stake tokens as well until a redeploy on 11 September 2026 removed that source, so that a single transaction can no longer take a position and mint the record that prices its exit. Caper describes the gate and the 21 September check.