KIP-88 DAO Guidance to the Cooperative

[clesaege]

KIP - 88: DAO Guidance to the Cooperative

Context

When Kleros was initially created, DAO tooling was almost non-existent, so a cooperative entity was created to conclude token sales, hold the treasury, develop the protocol and find integrations. Now the Kleros DAO is way more mature, we have better tools to evaluate proposal impacts (futarchy) and there is a desire within the community to be more involved.

This will of the community can be seen with the failure of “KIP-81: Strategic PNK Sale to Kleros Cooperative” and the numerous requests from users on telegram to give more power to the DAO.

We therefore propose that the Kleros DAO would be able to provide instructions to the Cooperative. Of course, the cooperative is never legally obliged to follow the instructions of the DAO, but we think that it can be pushed to do so with two levers:

• If the cooperative were to significantly disregard the DAO wishes, the DAO would commit to stop the PNK grants to the cooperative (and possibly give those to another entity or decide of funding directly).

• There should be some goodwill of the cooperative to follow community wishes. We can expect that a significant number of Cooperative members are also PNK holders and they could vote in the cooperative assembly to uphold the DAO wishes.

Proposal

We propose for the DAO to make requests to determine large cooperative orientations. Those requests should account for around 70-80% of the cooperative net spending. The remaining 20-30% being left for fast or discretionary decision making.

The DAO commits that if those requests are mainly fulfilled, the DAO will provide more PNK grants to the Cooperative in the future.
The DAO also commits that if those requests are mainly not fulfilled, the DAO will stop further PNK grants to the Cooperative and either fund one or multiple entities or spend its treasury directly.
In case it is not obvious whether those requests are mainly fulfilled or mainly non fulfilled, this proposal shall have no effect.

By itself, the proposal has no direct execution and establishes a governance framework. Further proposals would consist of requests to the Cooperative.

[n1c01a5]

Aligned on the spirit, but not on most of the specifics, which feel overall quite vague to me.

A few points where I see real gaps:

• No diagnostic. The proposal sets an abstract framework without documenting what’s actually wrong today (declining disputes, drift toward internal products, rising costs, transparency gaps, lack of community input on strategic decisions). Without diagnostic, it’s hard for the community to assess what “requests” the DAO should even make.
• No transparency requirements. Nothing on salaries, spending by project, or treasury reporting. The DAO would be asked to evaluate compliance without the data needed to do so.
• The 70–80% threshold is too loose. Who decides what falls inside the 70–80%? Who decides what counts as “mainly fulfilled”? The clause “if it is not obvious, this proposal shall have no effect” creates a built-in escape hatch.
• No explicit scope. Leadership, hires, project launches, the framework should name what the DAO is expected to weigh in on, not leave it fully open-ended.
• No public traceability on Cooperative member votes. The reputational lever you mention only works if the votes are actually visible.
• PNK grants should pause now. The Cooperative already holds significant ETH reserves. Continuing to mint PNK, diluting holders, while sitting on ETH that could fund operations doesn’t make sense. The Coop should spend its ETH treasury first but ultimately, whether to mint PNK or use ETH is a DAO decision.

I’ve drafted a more detailed version in parallel that covers these points: [draft] KIP-88|89: DAO Guidance to the Cooperative - Google Docs

[clesaege]

@n1c01a5

No diagnostic. The proposal sets an abstract framework without documenting what’s actually wrong today (declining disputes, drift toward internal products, rising costs, transparency gaps, lack of community input on strategic decisions). Without diagnostic, it’s hard for the community to assess what “requests” the DAO should even make.

You can put a “diagnostic” in comments. If the “diagnostic” put in the proposal is too opinionated, it will lead to people rejecting the entire proposal (because they don’t want to support the “diagnostic”) despite agreeing with a plan of action.

@n1c01a5

No transparency requirements. Nothing on salaries, spending by project, or treasury reporting. The DAO would be asked to evaluate compliance without the data needed to do so.

This can be done in another proposal. Good proposals should be like good code, do one thing but do it very well. If we end up mixing guidance and what the DAO would want to do with this guidance, some people will reject it because they disagree on what to do with this guidance despite agreeing that guidance is necessary.

@n1c01a5

The 70–80% threshold is too loose. Who decides what falls inside the 70–80%? Who decides what counts as “mainly fulfilled”? The clause “if it is not obvious, this proposal shall have no effect” creates a built-in escape hatch.

I think it is good because:

• We do not want the coop to be fully controlled by the DAO due to legal reasons. DAO should be client, not a controller.
• If there is full misalignment between the DAO and management, that could lead to drama (as some people would have “nothing to lose”. Keeping some control to management ensure that it still have residual power no matter what.
• There are times where we need to be opinionated and fast if we want to take opportunities.
• Not all the treasury come from PNK grants, a huge part come from government grants and fruitful investments by the Coop management.

@n1c01a5

No explicit scope. Leadership, hires, project launches, the framework should name what the DAO is expected to weigh in on, not leave it fully open-ended.

Left to other proposal, so that this one acts a minimum agreed framework. Other points should be discussed in individual proposals.

@n1c01a5

No public traceability on Cooperative member votes. The reputational lever you mention only works if the votes are actually visible.

Votes (outside of elections) are public by default. The assembly can ask for secret votes.
This doesn’t seem to be in power of the DAO to change that.

@n1c01a5

PNK grants should pause now. The Cooperative already holds significant ETH reserves. Continuing to mint PNK, diluting holders, while sitting on ETH that could fund operations doesn’t make sense. The Coop should spend its ETH treasury first but ultimately, whether to mint PNK or use ETH is a DAO decision.

Those grants are already paused for quite sometime, but obviously if price goes way higher it would make sense to provide more funding.

@n1c01a5

I’ve drafted a more detailed version in parallel that covers these points: [draft] KIP-88|89: DAO Guidance to the Cooperative - Google Docs

I’d appreciate if you could assign its own number to avoid confusions between the two proposals which do quite different things.

Overall I feel that you mainly agree on the proposal, but want extra things. Those extra things can be their own proposals. That is what this proposal does, open the gate for other guidance proposals.

[greenlucid]

Could it be legal from Coop standpoint to have DAO representation?

Here are some different ways it could be defined, numbers are reference points

DAO College, with 1 seat, and 20% of voting power

Whenever there are Coop meetings, since all meeting need warning with a certain amount of days, it gives enough time to formulate the question to the DAO, and create a Snapshot vote to establish the DAO’s vote on the topic. These votes are launched on any meeting, be it Extraordinary or Regular.

DAO Representatives College, with 3 seats, and 25% of voting power

DAO elects representatives for its interests on the DAO, who decide discretionally and are elected every 12 months via Snapshot vote, similar to representative democracy, and would act as a counter force to Coop workers and Coop leaders, as some kind of DAO Mission Board. But, these representatives could have their own agenda, and would be replaced if not acting towards the interests of the DAO.

DAO not represented officially, but voluntarily, some voters, making up 28% of the Cooperative vote, delegate to the DAO

A group of users, workers and maybe leaders, decide privately and voluntarily, or maybe according to extrinsic incentives, to delegate their vote to whatever the DAO decides in Snapshot votes. The DAO does not legally have representation, avoiding problems with the Cooperative legal situation.

[n1c01a5]

Beyond the framework itself, it would be valuable to share ASAP:

• A detailed breakdown of Cooperative spending over recent quarters (by project, by category, role)
• A performance review of the past years (what worked, what didn’t, what the lessons are)

Without that baseline, the DAO has no reference point to formulate meaningful requests and there’s a real risk of repeating the same patterns under a new label. A clear-eyed retrospective is the precondition for moving forward.

[n1c01a5]

On the Context: the framing here doesn’t fully reflect what’s actually pushing this proposal forward.

When Kleros was initially created, DAO tooling was almost non-existent, so a cooperative entity was created to conclude token sales, hold the treasury, develop the protocol and find integrations. Now the Kleros DAO is way more mature, we have better tools to evaluate proposal impacts (futarchy) and there is a desire within the community to be more involved.

DAO tooling has been mature for years the shift could have happened a long time ago and didn’t. What’s changed isn’t tooling availability, it’s that Kleros has been underperforming for years and PNK increasingly looks like a dying project to the market. The community no longer follows, complaints are accumulating, and engagement is declining across all channels.

To be fair, the community’s will is acknowledged in the next sentence (KIP-81 failure, Telegram requests). But it’s placed as a secondary signal after the “DAO is mature now” framing which inverts the real hierarchy of causes.

This is what’s bringing us here, not improved futarchy tools. It’s a political framing rather than a realistic one understandable from a “don’t offend anyone” perspective, but it sidesteps the actual diagnosis the community is asking for.

The honest version is closer to: the Cooperative has not delivered on its mission for years, the community has lost confidence, and the DAO needs a structural say to restore alignment between funding and outcomes. That’s the foundation a real governance framework should be built on.

[clesaege]

@greenlucid

DAO Representatives College, with 3 seats, and 25% of voting power

DAO elects representatives for its interests on the DAO, who decide discretionally and are elected every 12 months via Snapshot vote, similar to representative democracy, and would act as a counter force to Coop workers and Coop leaders, as some kind of DAO Mission Board. But, these representatives could have their own agenda, and would be replaced if not acting towards the interests of the DAO.

This is in line with the initial college vision (split the user college into jurors and apps using Kleros, in this case it would just be KlerosDAO using Kleros).
We could have an dapp college where dapps are given delegate slots in proportion of their importance in the Kleros ecosystem.
I think those representatives should:

• Always follow the DAO recommendations if there are any (or get replaced ASAP if they betrays).
• Vote according for their beliefs when the DAO is silent.

Note that this would be its own proposal (that both the DAO and the Coop should ratify).
Maybe @greenlucid you should start a separate discussion for this.

[KlerosJuror]

Push my proposal to vote please KIP-89, thanks.

[n1c01a5]

not aligned on the form, but aligned on the substance

I vote for

[FedericoAst]

There are some interesting points. I will try to add something to the discussion next week.

[clesaege]

The proposal has been put to vote.
Related futarchy market (here not required, but people can delegated to futarchist.eth to vote according to futarchy result).

[jnptzl]

voting No on this.

I don’t think the DAO is mature enough yet to take decisions of this scope. less than 10% of PNK has voted on this so far. we have a couple of loud voices on Telegram and fewer than 10 people giving opinions on the forum. we also have no known delegates outside the Cooperative who are committed to continued participation and deliberation. what we have are occasional contributions, with no follow-ups. the DAO members outside the Coop doesn’t hold regular meetings or do the work of finding common ground and forming an educated voice. without that, I fear most participants do not have a clear picture of what has been built over the last seven years, where the protocol sits commercially, or what running Kleros actually involves on a daily basis.

there is also no shared framework for evaluating the project, so every proposal gets debated on its own terms with shifting criteria. that’s fine when the stakes are opinions, but not fine when the stakes are 70-80% of cooperative net spending.

what I think happens in practice if this passes is that the Cooperative ends up writing the proposals the DAO then “instructs” the Cooperative to execute.

we would also be taking on governance debt, commit to a structure the non-coop members won’t actually run, then spend time patching it or quietly ignoring it.

also: the enforcement lever is too binary. cutting PNK grants to the Coop is a nuclear option, if it’s never used and the framework has no teeth, or it’s used once and we have a crisis in the coop.

I’d rather take the slower path, and have people to agree on how we should evaluate the project, onboard serious DAO participants that are committed more long term. hold more meetings between Coop and non-coop holders, where all of this is discussed more than once per year.

I do want the DAO to get to a point where we can trust it with more and more decisions, so I agree with the direction, but I think right now it’s premature and the DAO should earn that.

[FedericoAst]

Ok, finally I had some time to address this.

In general, I support the spirit of more decentralization and having more involvement from the DAO.

For me, the long-term vision was always that the DAO would progressively take more functions currently in the hands of the Cooperative.

But it’s also true my own thinking on this has shifted over time. Especially after seeing so many failed attempts of decentralization. Many projects that went furthest in the direction of putting more decisions in their DAO are returning to more traditional structures. I don’t think that means DAOs have no role. But I think maybe the right model is probably a hybrid that should be designed carefully.

In the case of Kleros, the Cooperative is a legal entity that already gives members many participation rights (including voting on major decisions, electing leadership and shaping strategic orientation). These rights don’t exist in many Web3 projects organized as foundations or other types of companies. Community members who want to propose changes in strategic direction or leadership, they can join the Cooperative and make a proposal there.

This said, I think the proposal has a number of problematic elements.

First, the feasibility. The Cooperative is incorporated as a legal entity under French law. Its decisions cannot be overridden by external instructions, regardless of their source.

The proposal seeks to address the enforceability limits through a sort of incentive mechanism conditioning “PNK grants” if the Cooperative complies with its “spending guidelines”. But this is not a binding mechanism. The Cooperative could very well comply and then DAO members change their mind and refuse to give that grant.

Moreover, the fact that the proposal speaks of a “grant” seems to imply there will be a future minting of tokens (from which I imagine the grant would come from). This would in itself be a controversial proposal not likely to pass (at the moment I’m opposed to minting more tokens).

Second, the current governance structure of the Kleros DAO is also problematic.

The governance is a rudimentary implementation of anonymous token-based voting without quorum requirements. This model carries well-known risks such as concentration of voting power into a few hands. If we add to this the fact of low participation which tends to happen in most DAOs (especially in a context of voting fatigue and bear market) this results in a scenario where a few whales (including myself, some would say) that could control decisions with a very small amount of tokens.

If the goal of this proposal is to increase governance decentralization, this is the exact opposite way of achieving it.

I’m very much in favor of a more decentralized governance, even including some input from the DAO. But this would require a way more careful design process that requires a reform in the governance of the DAO and its connection with the Cooperative.

In its current form, the value of this proposal is no better than just an “opinion poll” (and not even a good one). I’m not against doing polls in the DAO, but they should be done in a more representative way and without claims of imposing decisions on the Cooperative (again, an entity which already has its own decision-making mechanisms).

If people want to go the decentralization route, I think it should include the following items (in this order):

- An upgrade of the current DAO governance into a more representative system not based only on token voting and also including quorum requirements (maybe including elements of 1p1v, quadratic voting, separation of powers, etc.). There’s many examples of communities who have put a lot of thought into these issues to build more sophisticated governance systems such as Arbitrum, Optimism, Cardano and more.

Many of the ideas can be taken from the Blockchain Governance Toolkit developed by BlockchainGov, where I participated as an advisor.

The Blockchain Governance Toolkit | A Cookbook

- The feasibility issue could (at least in principle) be addressed by some legal engineering mechanism in the connection between the DAO and the Cooperative for the votes held in the DAO to have some binding element. This should be done in the context of mechanisms provided by French cooperative law. This could include a college where representatives are legally bound to vote in the way instructed by the DAO.

I think the solution could also include different Kleros products and mechanisms such as curation of proposals, SafeSnap and others. This also has the advantage of being a showcase of how other projects could implement good governance systems by bridging the on-chain and off-chain governance. It would be a way to integrate Kleros products into more DAOs.

This would embed DAO influence directly into the Cooperative’s governance structure rather than creating an awkward “external instruction channel”. I think this would be a way to build a more durable and long term solution building on what the Cooperative structure already offers.

For these reasons, I’m against this proposal.

[swanherder]

This is interesting and largely seems to make sense

[n1c01a5]

First, the feasibility. The Cooperative is incorporated as a legal entity under French law. Its decisions cannot be overridden by external instructions, regardless of their source.

The Cooperative is free to choose whatever governance it wants and nothing prevents it from choosing to follow the DAO’s guidance. That’s exactly the point.

As the proposal says so explicitly: the Cooperative is never legally obliged, the mechanism is incentive-based, not coercive. An incentive doesn’t need to be legally binding to work.

The proposal seeks to address the enforceability limits through a sort of incentive mechanism conditioning “PNK grants” if the Cooperative complies with its “spending guidelines”. But this is not a binding mechanism. The Cooperative could very well comply and then DAO members change their mind and refuse to give that grant.

Moreover, the fact that the proposal speaks of a “grant” seems to imply there will be a future minting of tokens (from which I imagine the grant would come from). This would in itself be a controversial proposal not likely to pass (at the moment I’m opposed to minting more tokens).

You’re right that “grant” is ambiguous. I had flagged it too, and Clément also clarified that this does not involve minting new tokens (at least in the short-to-medium term). The wording could have been clearer, but the substance isn’t about minting.

Second, the current governance structure of the Kleros DAO is also problematic.

The governance is a rudimentary implementation of anonymous token-based voting without quorum requirements. This model carries well-known risks such as concentration of voting power into a few hands. If we add to this the fact of low participation which tends to happen in most DAOs (especially in a context of voting fatigue and bear market) this results in a scenario where a few whales (including myself, some would say) that could control decisions with a very small amount of tokens.

If the goal of this proposal is to increase governance decentralization, this is the exact opposite way of achieving it.

Maybe, but it would still be more decentralized than the Cooperative which is itself run by a far narrower college. Imperfect token voting is a wider base than the status quo, not a smaller one.

In its current form, the value of this proposal is no better than just an “opinion poll” (and not even a good one). I’m not against doing polls in the DAO, but they should be done in a more representative way and without claims of imposing decisions on the Cooperative (again, an entity which already has its own decision-making mechanisms).

If you genuinely support decentralization, this is not “just an opinion poll”.

If people want to go the decentralization route, I think it should include the following items (in this order):

- An upgrade of the current DAO governance into a more representative system not based only on token voting and also including quorum requirements (maybe including elements of 1p1v, quadratic voting, separation of powers, etc.). There’s many examples of communities who have put a lot of thought into these issues to build more sophisticated governance systems such as Arbitrum, Optimism, Cardano and more.

A multi-college reform under French cooperative law would take months or years. In the meantime, what I had proposed: quadratic voting + Proof of Humanity would already guard against the whale problem you raise, without blocking all progress until the full reform lands.

Finally, I’d like to add about

In general, I support the spirit of more decentralization and having more involvement from the DAO.

I don’t think so:

• Asking the DAO for input is minimal and rarely strategic: only 40 proposals in almost 7 years, most of them technical (court parameters, etc.) rather than strategic direction.

• Development is closed: the main repo has 3 maintainers, and over the last 3 months contributions come from a handful of internal contractors only and, to be verified, no external contributions.

  

• Communication is partisan: the pinned tweet is about Argentina, but there wasn’t a single tweet about KIP-88, a major governance proposal. The first posts on the Kleros account are dominated by you personally (multiple posts + videos of yourself). The official channel amplifies personal content but stays silent on a flagship DAO proposal (there is a tweet about the kip).

  

• Project funding raises conflict-of-interest questions: some projects are funded without community consultation, possibly at the expense of others.

• Transparency is lacking: I’ve asked several times for a breakdown of the Cooperative’s spending and never got an answer.

• The weak PNK is the result of strategic choices: certain reward designs, drift from the original vision, and a lack of partnerships have weakened the token which is precisely what creates the concentration risk you point to.