Updated v1.4.313 revisions
Cite
Caper. "Aragon." October 13, 2026. https://caper.network/wiki/dao-governance/tooling/frameworks/aragonAragon is one of the oldest and most influential DAO frameworks – infrastructure for deploying an organization whose rules, membership and treasury live entirely on-chain. Founded in 2016 and funded by a $25m token sale in 2017 (Aragon), it let non-programmers spin up a DAO with token voting, a treasury and upgradeable permissions, and it underpinned the early on-chain governance of major protocols including Curve and Lido. Its more recent history is also a cautionary tale about token-governed foundations, summarised below and told in full on Aragon DAO. (Aragon)
The original stack – aragonOS plus the Aragon client – modelled a DAO as a set of on-chain apps (voting, tokens, finance, permissions) wired together through a central access-control list. Its defining idea was upgradeable, granular permissions: every action a DAO could take was a permission that could be granted to an address, a token vote, or another app, so an organization's power structure was itself programmable. This made Aragon a natural home for protocol DAOs that needed to hand contract control to a token vote.
In 2023 Aragon shipped a ground-up rewrite, Aragon OSx, built around a lean permission-management core and a plugin architecture: governance mechanisms (token voting, multisig, optimistic, admin) are installed as swappable plugins rather than baked in, so a DAO can evolve its own rules over time. OSx is deployed on Ethereum, Polygon and other EVM chains and is the framework Aragon actively maintains today. (Aragon: A New Chapter)
In November 2023 the Aragon Association announced it would dissolve, wind down token-based governance of its treasury, and refocus as a product-oriented team building OSx. Rather than transfer the treasury to a DAO, it returned it to token holders: an autonomous redemption contract, funded with 86,343 ETH, let ANT be redeemed at a fixed 0.0025376 ETH per ANT through 2 November 2024, after which unredeemed ANT was sent to a burn address and remaining ETH passed to the successor structure. Burnt there means transferred rather than destroyed: read on 19 September 2026, the token contract still reports a total supply of 43,192,488 ANT, of which 36,922,472, or 85.5%, sits at the 0x…dEaD address, and the redemption portal still answers.
That decision followed an eighteen-month dispute over a treasury transfer ANT holders had already voted for, which the association halted in May 2023 after activist buyers assembled a majority – the sequence, and both sides' accounts of it, are documented on Aragon DAO. Of the ANT in circulation, 87% was ultimately redeemed across just 2,670 redemptions, paying out 75,093.99 ETH (Aragon's final report). It is one of the most-cited examples of a major project concluding that a large token-governed foundation treasury was more liability than asset. (A New Chapter for the Aragon Project)
Describing Aragon as the maintainer of a DAO framework understates where the product went after the wind-down. Its work since 2025 has concentrated on governance that sits inside other protocols' security models and token economies rather than on standalone DAO creation. On 30 June 2025 the current Aragon app replaced every earlier version; existing DAOs kept working unchanged, because their proposals and settings live in OSx contracts (Aragon).
Guardians. Aragon Guardians is a veto layer built on OSx: a protective role that lets a community or a set of trusted actors block a malicious proposal, an exploit or a buggy upgrade before it lands. It is used by MetaMorpho curators and by DeFi protocols including Steakhouse Financial, Hyperithm, MEV Capital and Re7 Labs. In October 2025 the Ethereum Foundation disclosed that it holds 2,400 ETH and roughly $6M in stablecoins in Morpho yield-bearing vaults safeguarded by Aragon Guardians and the Aragon TokenVoting plugin.
LockToVote and Guardians v2. First used for Guardians v2 in Morpho Vaults v2 in September 2025 (Aragon) and announced as a plugin release on 4 November 2025 (Aragon), the LockToVote plugin lets any ERC-20 be used for governance without being a purpose-built governance token: holders lock tokens into a LockManager contract, the locked balance becomes voting power, locked tokens are non-transferable so the same balance cannot vote twice from two addresses, and unlocking resets voting power to zero. The design keeps tokens liquid for trading or collateral right up to the moment a decision is live, and gives vault depositors a real-time veto over curator actions. In October 2025 Aragon said the plugin was available to whitelisted projects only (Aragon); the November release post does not repeat that restriction.
Vote-escrow and gauges. Much of Aragon's 2025 work was tokenomics rather than voting: a modular vote-escrow library, sold as its Value Accrual Toolkit, in which holders lock tokens for voting power and direct emissions through gauges. Mode, Katana, YieldBasis and Celo's public-goods gauges run on it; how their parameters differ from Curve's original is on vote escrow. In January 2026 Aragon added tokenomics advisory to what it sells (Aragon), and on 4 February 2026 it launched the Ownership Token Framework, a public dashboard that scores what a token's holders control; see token valuation.
Treasury, distribution and permissions. The Capital Distributor (September 2025) computes who may claim what and what a claim triggers (airdrops and token distribution). Linked Accounts (March 2026) join separately governed accounts into one organisation view (DAO treasury management). Permissions-based access control (August 2025) limits each governance process to named contract functions (typed vs arbitrary execution), and the app's governance designer can give a role such as veto to any EVM address, including a Safe or another DAO (Aragon, May 2025). Taiko's DAO, launched with Aragon in January 2025, is built that way: a security council proposes and token holders veto (security councils). BORGs, added in September 2025 with MetaLeX, wrap chosen addresses and roles in a legal entity (DAO legal structures).
Policy plugins. In January 2026 Aragon set out a direction it summarises as "the art of not governing": a class of plugins that, unlike governance plugins, execute without requiring input, applying a rule configured once to a bounded and known action type. The argument is that proposal-based governance treats every action as an onchain referendum, and that recurring operations such as payroll, vendor top-ups and rebalancing do not need human intent re-encoded each time (Beyond Proposals Pt. I). The cost side of that argument is on who pays to execute a passed proposal. The same idea, read at contract source in Lido, Maker, Aave and Compound, is on standing authorizations.
Private voting. Aragon has three designs for secret ballots, and each trusts a different party. All three postdate Vocdoni, the anonymous-voting team the Aragon Association bought in January 2021, which now presents itself as a Swiss non-profit association beside a for-profit company. A MACI plugin, built with Privacy & Scaling Explorations and announced on 20 August 2025, relies on a coordinator who can decrypt every ballot, and Aragon offered it as a custom build rather than a standard plugin (Aragon). CRISP, from Gnosis Guild's Enclave project, since renamed the Interfold, splits decryption across a committee of independent nodes; Aragon announced the integration on 10 November 2025 (Aragon) and published an OSx plugin on 30 June 2026 that runs on testnet, with production waiting on the Interfold network's rollout (Aragon). Confidential Voting, below, is the one released as a standard plugin.
Confidential voting. On 10 September 2026 Aragon released Confidential Voting, an OSx governance plugin built on the Zama Protocol’s fully homomorphic encryption. A ballot is encrypted in the voter’s client and stays encrypted while the plugin checks that it is exactly one of Yes, No or Abstain, applies the voter’s weight and adds it into encrypted totals. Only the three totals are decrypted, by threshold decryption after the voting period ends, and support and quorum are evaluated against them.
By default nobody holds decryption rights over an individual ballot, though an organisation can grant a designated party such as an auditor access under set conditions. The design rules out a running tally and early execution, and it lets a voter replace an encrypted vote until close, which Aragon presents as coercion resistance: a vote shown to someone before the deadline does not prove the final choice. The properties it does and does not provide are compared with Snapshot’s and MACI’s on private and shielded voting.
Verification and identity tooling. Two June 2026 releases sit beside the governance plugins rather than inside them. EVM Mirror (15 June 2026) addresses the gap between the Git commit an auditor read and the bytecode users transact with: it compares a block explorer's verified source against a checkout, package or artifact the reader already trusts. Aragon built it because it deploys across more than fifteen chains and sits as an independent signer on other protocols' security councils, among them Spark's and Taiko's (Aragon, January 2026), and it names Lido's DiffyScan as the prior art it outgrew. Onchain Profiles (5 June 2026) attaches a public identity to a governance address. The verification problem is the same one this wiki treats under DAO metrics and analytics: an address is not a deployment, and a deployment is not the code that was reviewed.
Aragon's lasting contribution is architectural: the notion that a DAO is a permission system with pluggable governance, and that those permissions should be upgradeable, runs through much of the tooling that followed. As a live product it is no longer the token-governed network it once was, a distinction the wiki keeps clear because older third-party descriptions still present the ANT-governed Aragon Network that no longer exists. What replaced it is narrower and more specific than "a DAO framework": OSx as a neutral execution environment, and a plugin catalogue sold into protocols that need a governance or veto layer around assets they already hold.
Not every customer stays: Geo, a knowledge-graph app whose community spaces launched as Aragon DAOs in January 2025, specified a replacement contract system in October 2025 that removes Aragon OSx entirely, citing the complexity it added (Geo). Both aragon.org and the OSx app are live. The blog published six times between January and June 2026, then not again until 10 September 2026, when it announced Confidential Voting (feed read 18 September 2026).
The code tells a narrower story than the announcements, and 2026 changed how it is distributed. On 18 May 2026 Aragon archived osx-commons, the repository that had held the contracts shared between OSx and its plugins, and folded its source into the main osx repository in a refactor that also moved the build from Hardhat to Foundry. The changelog for that release states that the @aragon/osx npm package is retired and that the Solidity is now consumed as a git submodule, and it adds one core component, a MemberRegistry in which an address claims an ENS subdomain under a parent the organisation configures. The same entry pins ProtocolVersion at 1.4.0, so nothing in the governed surface changed (Aragon).
Two numbers set the scale of what is deployed and what is still depended on. The published address set carries a DAOFactory on nineteen mainnets plus two testnets (Aragon). And the JavaScript SDK, archived since July 2024 and marked on npm as no longer supported, was still downloaded 1,288 times in the month to 18 September 2026 (npm) – integrators building against a client its own authors have retired. The newest plugin repository, created on 27 July 2026, is a cross-chain controller that lets a DAO execute actions on another chain.
Aragon is a general framework you assemble a DAO from; Caper is a purpose-built protocol with the governance, treasury and exit rules already fixed. The two sit at opposite ends of a real trade-off – Aragon's plugins maximize configurability, Caper's fixed rules minimize the surface a caper's members have to trust or re-litigate. Aragon's Association wind-down is also a pointed argument for Caper's stance that treasury claims should be an on-chain member right (redeemable at exit) rather than something a foundation decides how to return.