With almost 200 jurors onboarded to Aragon Court, whose first term starts on February 10th, we want to present Aragon One’s plans for how Aragon Court will transform the experience of using Aragon organizations with Aragon Agreements.

Aragon Agreements leverage Aragon Court to allow Aragon organizations to benefit from the richness of subjective terms and conditions that can either not be expressed in smart contracts or when doing so would result in a complicated and slow-moving organization.

Agreements are a generalization of the previously described Proposal Agreements, that would have allowed organizations to have subjective rules for what types of proposals were allowed in the organization.

Agreements take it a step further by allowing subjective text to govern any action someone can perform in an organization, including but not limited to proposals.

Managing a traditional entity can be done optimistically, because there's usually social capital at stake or, at the very least, trust in the institutions that offer recourse if someone acts in an illicit way like calling up the bank to revert a transaction or taking someone to court.

For example, even though multiple people in a company may have access (technically a 'permission') to perform transfers out of the company's bank account, there is minimal risk that someone will illicitly transfer money because of the legal and social consequences. These deterrents prevent bad behavior so well that administrators can trust authorized members to make transactions autonomously while making the organization more fluid and efficient.

On the other hand, DAO management must be done pessimistically, and for good reasons. When dealing with open organizations with pseudo-anonymous members, there's no social capital involved (a refrigerator can control an account with DAO tokens), and there's little to no recourse for malfeasance (transactions usually cannot be reverted, and legal recourse is close to impossible).

Agreements are a way of giving members special permissions that make the experience of using an Aragon organization akin to managing a traditional entity, while still keeping the advantages of sovereign digital organizations.

Organizations will be able to create different agreements for different actor groups (for example, token holders submitting proposals, members acting as executives, technical members that can upgrade smart contracts). Given Aragon Court’s subscription model, an organization will need to pay Aragon Court a small flat fee per the agreement, purchasing the right to use Aragon Court to dispute whether actions should be permitted. This enables a secure Court, in which jurors earn decent returns for simply being available to work, even if organizations rarely create disputes because actors know that the Court will resolve them fairly (and there’s no incentive to misbehave).

Agreements have the potential to convert crypto-native organizations (no legal wrapping, no KYC) from pessimistic organizations with massive overhead for every operation into optimistic organizations that operate with efficiency while protecting stakeholder interests as collectively defined in their Aragon Agreement.

Aragon Court will serve as the deterrent for violating agreements and encourage actions clearly allowed by the agreements.

A new interaction model for DAOs

A common characteristic of all decentralized governance experiments to date, and a popular target for detractors, is low voter turnout. The nascent state of crypto's user experience and seemingly endless votes required to approve every mundane action makes voting in DAOs far from a delightful experience.

These frictions have led to a common pattern in DAO operations where there's a person or small group that pushes the organization forward by instigating most of the proposals and whipping votes from token holders to execute actions. We can see this pattern in the most successful DAOs to date, with Moloch's Ameen, Saint Fame's Jacob, or LexDAO's Ross.

This pattern isn't necessarily bad since they don't explicitly possess any special powers, and any other member could step up and try to steer the organization in a different direction. Agreements could allow these "soft" leaders to perform low stake or reasonably uncontentious actions in the organization without the overhead of needing a majority vote from all token holders.

Aragon Agreements open the door to a brand new interaction model for DAOs. Organization stakeholders make less frequent but more critical decisions by focusing on setting the direction and rules but allowing engaged actors to execute actions on behalf of the DAO more efficiently. Stakeholders' valuable attention is no longer required in the day-to-day.

With Agreements, stakeholders keep their sovereignty while dramatically reducing overhead for the long tail of actions the DAO executes.

Collateralized delayed execution

At a high level, an Aragon Agreement is a set of non-computable rules and conditions that define whether organizational interactions are licit or illicit. Given that a smart contract cannot verify these conditions, it requires human supervision to flag illicit actions and a dispute resolution mechanism (Aragon Court) to make the ultimate decision of whether an action should be permitted and to penalize bad actors.

When creating an agreement the organization can decide the set of actions that agreement signers will be able to perform, the subjective rules for how to perform those actions and the security parameters for execution, like a time delay before execution, how much time after submitting an illicit action someone can receive a penalty and how much collateral signers must deposit to enter.

The organization also decides who can sign the agreement in the first place. This can be done with maximum flexibility by having an agreement that is open for anyone to sign, or a limited whitelist of accounts and entities holding a certain amount of organization tokens. In terms of user experience, most will only interact with the agreement at the time of signing and terminating the agreement, unless an action they perform is challenged. Once they sign the agreement, they will be able to use the organization as before, and unless they are performing a high volume of actions, they shouldn't need to add collateral frequently.

When someone signs an agreement, they are automatically granted a set of permissions they can use to perform during the time that the agreement bounds them (they have enough collateral to back their actions) and the conditions required to sign the agreement in the first place still hold (i.e. owning a certain amount of organization tokens).

Agreement signers will then be able to submit actions into their queue and schedule them consecutively for execution after the security delay time passes.

During the time  transactions are in the queue, observers can challenge actions they believe violate the signed agreement. At this point, there will be a chance for the proposer and challenger to settle on a penalty and whether the action should be executed or not. If they do not reach a settlement in a certain amount of time, the dispute is escalated to Aragon Court, which decides whether the action should execute or not and who receives a penalty.

Even if an action has already been executed, and therefore unlikely or impossible to be reverted, signers can be penalized for their actions by either reaching a settlement with a challenger or a Court ruling. This allows having very short or no delay time for trusted entities, while still having some recourse if they perform a malicious action.

Stakeholder sovereignty in contentious times

Even though the long tail of actions an organization performs are fairly uncontentious, a small minority of them are indeed contentious and have a big impact on the future and direction of the organization.

These decisions must not be made a single actor, even if the action is technically legal per the agreement.

Agreements allow an entity with permissions (i.e. Voting instance) to veto any action in the queue and partially slash whoever submitted it, without taking the action to Court.

This allows stakeholders to maintain their sovereignty by influencing contentious decisions that will shape the organization, while at the same time not having to make decisions for small daily operations.

Beyond delays, taking governance fully off-chain

Agreements can be taken one step further and allow stakeholders to control their organization in a completely off-chain way, without having to use wallets, private keys, or signing messages or transactions for daily operations.

One can argue that governance is already occurring almost fully off-chain and quite rapidly. Ideation, debating, and whipping is already happening fully off-chain, and are the most important activities of any governance system.

The only part of the process that is happening on-chain is the tallying of votes and the execution of actions when approved.

Agreements take this even further by moving voting tallies off-chain. An organization could adopt norms for gathering consensus using the same tools where debate is already happening, be that Telegram polls, Discourse likes, or any other mechanism.

Once a member or an oracle sees that there's enough support for a certain action, they could submit it on-chain for execution. Other actors may challenge the action because that the decision didn't follow the specified process in the agreement or that there is not enough support to perform it, providing evidence to the Court in order to stop the action and slash the bad actor.

Gathering consensus directly from the tools where conversations and discussions are happening will dramatically improve user experience and allow organizations to move much faster than was ever possible before.

This post was a collaboration between

Jorge Izquierdo, Aragon One

  • Jorge Izquierdo
  • Aragon One