[SEP #4] SafeDAO Constitution

  • Title: SafeDAO Constitution
  • Authors: Lukas Schor (with early support from Longhash Team)
  • Created: 2023-01-10


This proposal introduces a constitution for SafeDAO, defining its stakeholders as well as its mission, goals and principles. It is proposed that changes to this constitution require a stricter governance process than regular SEPs, that is double the quorum and a ⅔ majority vote.

Proposal details

Purpose and Background

Decentralized autonomous organizations contain a particular promise: the ability to solve old coordination problems through ‘smart contracts’ that encode constitutional rights. Although smart contracts are promising new tools for governance, they alone cannot govern communities; traditional constitutions and declarations of rights are still crucial for good governance. (Metagov)

DAOs set out to improve coordination through technical means, but soon realized that a governance smart contract is not sufficient. Without a north-star to rally the community behind, a DAO can suffer from fundamental disagreements on its general direction and lack of guidance on how to evaluate proposals.

Introducing a constitution helps resolving these issues. By agreeing to a mission, goals and principles, the DAO has more certainty on its nature and ambitions. By defining a higher threshold for changes to the constitution compared to regular proposals, the DAO avoids fundamental changes through proposals that have not received sufficient support by the community.

Through the mission, goals and principles defined in this proposal, SafeDAO will be more aligned and operate more efficiently when introducing future proposals. Specifically, I suggest the following text as SafeDAO’s initial constitution:



SafeDAO is a decentralized collective, stewarding a thriving ecosystem around the smart contract wallet standard for Ethereum and other blockchains (“Safe Ecosystem”). SafeDAO is governed using the Safe Token (“SAFE”).


SafeDAO’s mission is to establish smart contract wallets as the default means for interacting with web3, also known as account abstraction. Current self-custodial wallets lack the security and convenience to be a viable alternative to custodial solutions for most people. In contrast to regular wallets (EOAs), smart contract wallets provide characteristics such as recovery mechanisms, granular access control, upgradability, batching transactions as well as hybrid-custody setups. By building web3 around smart contract wallets, we reduce the reliance of users on custodial solutions, reducing counterparty risks, walled gardens and potential for censorship.

As SafeDAO, we aspire to work towards this mission by establishing standards for composable smart contract wallets (the “Safe Protocol”). Through open source components and ecosystem initiatives we are nurturing the Safe Ecosystem, consisting of wallets, infrastructure, tooling and services. The Safe Ecosystem is building on a shared implementation of the Safe Protocol and is catering to the needs of various user-groups and use-cases in web3. SafeDAO promotes the common welfare, sustainable development, robustness and diversity of the Safe Ecosystem.


SafeDAO shall use the following goals to work towards its mission:

  1. Foster a vibrant ecosystem

SafeDAO benefits from a vibrant ecosystem built on shared components and standards. SafeDAO supports and empowers new and existing projects integrating with the Safe Protocol or initiatives supporting the usage of Safe Protocol.

  1. Resilience through decentralization

The Safe Ecosystem shall be independent from any single entity. The Safe Ecosystem components, including governance, should be decentralized and permissionless.

  1. Tokenize value

SAFE should represent the value of the Safe Ecosystem. Mechanisms should be created to link the growth of the Safe Ecosystem to the growth in utility of SAFE.


All proposals passing the SafeDAO governance process shall adhere to the following principles:

  1. Safe Protocol is a public good

The Safe Protocol is core infrastructure for web3. Any changes that may be introduced do not compromise its public good nature and carefully balance the interests of different stakeholders in the Safe Ecosystem. This incentivises the unity of the Safe Ecosystem and, as a result, maintains the property of Safe as a portable account.

  1. Token utility is aligned with ecosystem interests

Any utility of SAFE directly or indirectly adds value to the Safe Ecosystem. Changes to token mechanisms need to consider the effects on different stakeholder including user, builder and token holder.

  1. Use of treasury is limited to achieving SafeDAO goals

The usage of any assets associated with the SafeDAO treasury is credibly linked to one or more of the SafeDAO goals outlined in the SafeDAO Constitution.


Changes or amendments to the constitution require a two-thirds majority and double the usual Safe Ecosystem Proposal (“SEP”) quorum at any given time.

Effects and Impact Analysis

The intended main effect of this proposal is to give SafeDAO stakeholders an explicit way to coordinate around a shared mission and high-level means to get there. On a more practical level, this constitution document should have the effect of informing every stakeholder’s thinking, writing and decision-making in a way that is more likely to be in line with the general direction of SafeDAO, i.e. arguing and voting for proposals that adhere to the constitution and against if not. Assuming that the constitution in its current and future form will always be optimized for the success of the Safe Ecosystem, adhering to the constitution should have a direct impact on the success of the Safe Ecosystem.

An argument against this constitution and its higher governance threshold compared to a regular Safe Ecosystem Proposal (SEP) is the fact that SafeDAO will be more constrained in changes relating to its constitution. However, as mentioned before, this should be seen as a feature rather than a bug as it provides security and steadiness around a few fundamental elements of SafeDAO that are vital to its long-term success and should be protected by a governance threshold with higher difficulty than regular SEPs.

In terms of risks, this constitution could define a direction or high-level means that turn out to be less ideal or even counter-productive. This risk is mitigated by allowing changes to this constitution, as described in article V.

Another risk lies in this constitution being ignored, not taken seriously enough or maliciously stretching an interpretation of its contents to a point that is against the intent of the constitution. Parts of this risk can be mitigated through a dispute resolution mechanism to keep SafeDAO and its SEPs in check and provide a web3-native process to refuse SEPs that are against this constitution. Other parts can be mitigated through soft-enforcement such as requiring SEP authors to document who a proposal is related to the constitution and where there might be conflicts. Such enforcement mechanisms could be introduced through a future change in the governance mechanism of SafeDAO.

Alternative Solutions

  • No constitution: SafeDAO could choose not to introduce a constitution and proceed with regular SEPs only. In this case, the DAO would not have as much alignment on a general purpose and direction as it would have with a constitution, resulting in more information asymmetry and less coherently coordinated SEPs. Without a constitution, the DAO would also be at risk of being radically redirected through a regular proposal, whereas this constitution defines fundamental properties that can be changed, but not as easily as through a regular SEP.

  • A constitution-like document under another title: While many DAOs introduce constitutions, some adopt similar but distinct forms, such as community covenants, code of conducts, manifests or lores. Some of these options were disregarded due to their specificity, such as a code of conduct, whereas a constitution provides SafeDAO with a generalistic foundation. The title ‘constitution’ was prefered because it is universally understood globally as a foundational document.

  • A constitution with a different mission, goals and principles: While details in this constitution may well need to be changed at a later time, the overall approach to build an ecosystem around a use-case agnostic protocol remains key.

    • SafeDAO as a product DAO: The SafeDAO Constitution could be focused on Safe being a vertically integrated product, consisting of smart contracts, infrastructure and user interfaces, all being provided and maintained by the SafeDAO. Doing so without fostering an ecosystem is a short-sighted effort, as existing Web3 projects have showcased that building a thriving ecosystem around a protocol can be much more impactful than just focusing on creating a product or suite of products.
    • Safe as DAO tooling: One of the most visible use-cases for Safe is its use within the operations of other DAOs. Either as a Multisig solution for teams / working groups or as a DAO avatar via tools like Zodiac Reality. The SafeDAO Constitution could focus on this user-group and double-down on Safe as a DAO tool within the mission and goals. However, there are already a wide range of other user groups and use-cases currently driving value to the Safe Protocol. This includes institutional users, retail users, or progragramitc use-cases of Safe within a DevOps function, for example. The chosen framing of the SafeDAO Constitution recognises the use-case agnostic nature of the Safe Protocol and anticipates many more use-cases to emerge in the future, therefore not artificially limiting the scope of the mission.

Technical Implementation

This proposal requires a new Snapshot Subspace to the main Snapshot space safe.eth for SEPs. The new Subspace settings reflect double the quorum and a ⅔ majority to pass a vote.

Furthermore, the SafeDAO Constitution should be highlighted prominently across SafeDAO’s documentation, communications channels and onboarding processes to ensure that every current and future SafeDAO stakeholder is sufficiently aware.


Copyright and related rights waived via CC0.


Good Proposal, I support this.

1 Like

This is an excellent proposal and provides much needed clarity regarding SafeDAO direction and purpose. I support it.


I support this constitution. Thanks for putting it together!

Would you clarify this point?

Does this mean no fees can be transferred directly to investors? Is it acceptable to route fees to a treasury?

It seems unnecessary to create a new Subspace to accommodate new quorum and majority rules given that the constitution is not yet binding. I suggest waiting to propose a new Snapshot Subspace and including that change in any future proposal that makes the constitution binding (through the use of Kleros Court, for example).

1 Like

This is an interesting discussion of what defines a public good.

Seems to hint at a future where SAFE will introduce fees and look to achieve profitability. Don’t blame them either as it is one of the highest amounts of ETH secured in a smart contract.

The Technical Implementation sounds interesting. I have rarely seen Subspaces used in practice - I support the experimentation of this Space but would leave room for future amendment.

Introducing more load on Governance participants can be exhausting and deter participation - especially if you have a 2/3 quorum; that’s pretty high!

Feels like most change can come in these parameters.

In regards to the constitution, I am appreciative of its brevity and focus on utility / value for the $SAFE token.

Supportive of the general Goals and Principles of the constitution, especially one that expands beyond current use case around treasury management and take a more use-case agnostic/ pluralistic view.

In the near term I think it is too early for value extraction. A focus on being public good instead of being a product DAO will eliminate all the conflict of interests which stems from many delegates and guardians who are already building on the application/product layer

As others have asked, not sure why new Snapshot Subspace is a requirement here?


Thank you for creating this !

I got two suggestions

  1. To further reduce the risk of this constitution being ignored, can we create 1-2 examples of what would be considered a good proposal and what would be a poor proposal or something that might not align. Since one could argue even in a roundabout way that most proposals can indirectly benefit the Safe ecosystem.

  2. Since there is a document that outlines the north star and direction, can we tweak SEP process to be quicker with some more additional safe guards and revaluate after 6 months.


The current wording does no explicitly exclude the possibility of fees or describe where fees are being directed. That being said, I think it is critical for the success of Safe and SafeDAO that any fees do not compromise the public good nature of Safe Protocol. I.e. the token should capture value through utility within the ecosystem rather than pure value extraction.

To be honest, I’m a bit surprised by the backlash on the Sub-Space part. I don’t really see a big downside of using a sub-space to facilitate future changes of the constitution. If using the current parent-space, a change to the constitution could be displayed in Snapshot as “passed” while it technically failed, potentially causing confusion. A Sub-space seems like a low-effort way to prevent this.

Curious to hear what the concrete hesitations are on this? Happy to drop this part if there is tangible drawbacks, but as none of the comments have so far pointed out any, I’d first want to hear some more thoughts on this.


This is an excellent proposal and provides much needed clarity regarding SAFEDAO direction and purpose. I support it.


Thanks for the clarification.

We’re in agreement that we shouldn’t create a mismatch between the pass/fail as displayed on Snapshot and the proposal result as defined by the SafeDAO governance process. My belief is that there should not be separate quorum and majority rule thresholds for changes to the constitution at this point, especially since the constitution is not binding and functions more like a statement of purpose.

In the early stages, I’m cautious about defining a governance process that inhibits change, complicates governance, or has high quorum requirements because lack of governance participation is an issue in many DAOs.

SEP #1 reached 14M total votes and SEP #2 has 13M so far, so neither of them would meet the quorum requirement of 20M defined here. It would be comforting to see that we can reach 20M votes on SEPs before requiring that level of participation for certain changes.


Following up here since SEP #2 closed with 32M votes. :partying_face:

It looks like we can hit the 20M quorum, which is great. I still think we should keep the process simple and hold off on having two sets of proposal passage requirements until the constitution is more consequential–adding this context to qualify my previous post.


Thank you @lukas for taking the time and effort to communicate the principals and purpose of the SafeDAO! This captures the mission and direction the Safe protocol is moving towards. Potentially adding details/examples to the Goals and Principles sections could be beneficial when analyzing future SEPs and comparing their stated outcomes to the goals defined in this constitution.

Constitution quorum

The higher 20mm SAFE quorum seems important for the constitution because it defines the guiding values and mission of SafeDAO. The significance of these guiding principles outweigh the added complexity of a subspace.

Decentralized collectives

The idea of the “SafeDAO as a decentralized collective” makes sense as most “DAOs” are in actuality closer to a Distributed Collective (DISCO) defined by @castig of Console.xyz.

III. Goals

  1. Foster a vibrant ecosystem

It could be useful to describe metrics that are indicative of a vibrant ecosystem to help define and measure what a vibrant ecosystem is?


  • Assets under management (AUM) on Safe protocol
  • User activity on Safe and apps built on Safe protocol
  1. Resilience through decentralization

Same point as for 1.. How can we tell if Safe protocol is decentralized?


  • Can apps on the Safe ecosystem run without Gnosis infrastructure (potentially broken down by infrastructure components)
  • SAFE token distribution
    • Not voting delegation since users can choose to override their delegate if desired
    • Will remain constant in the short-term, but useful once SAFE is transferable
  • User distribution across Safe app ecosystem

It may be useful for these items and others that people contribute to be added to the constitution in order to help measure success.

IV. Principles

Any fees that may be introduced do not compromise its public good nature and carefully balance the interests of different stakeholders.

It may be useful to define who the stakeholders are so that way future SEPs can address the impact of specific stakeholders.


  • Users of safe apps
  • Core developers, business, and support team of Safe protocol(s)
  • SafeDAO Guardians
  • Developers building apps on-top of Safe protocol
  • SAFE token holders
  • Gnosis team

Thanks for starting this conversation. I feel much detail in missing In you’re proposals current state, ie, vague references and unclear reasonings.

for example…

SAFE does not require fees of any sort nor does it need them. Further details on why fees were referenced and elaboration on what “carefully balance the interests of different stakeholders” is referring to is required.

Aside from that comment, i find that this has set a good foundation to build on and expand upon. Notably the paragraph that mentioned the use of Kleros Court as being a way to dispute disagreements. Community forms, being as widely used as they are, is limited in capabilities and can be chaotic at times. looking for alternative ways to collaborate with community members is important to grow.

1 Like

This is something that I think we should touch on.

I brought up some uncertainty around the large voter addresses earlier in the week— and in another thread.

Here is the link for it: [SEP #2] Community Initiative To Unpause Token Contract (Enabling Transferability) - #244 by raynemang

If you have some time, can you take a look at my analysis? I’d love to get some of my findings cleared up or explained :smiley:

Thank you, @lukas!


Thanks for sharing these perspectives.

I saw this article in the website you shared

I think this article serves as a good guide for this proposal.
There is a very detailed explanation about the meaning of DAO and the meaning of the DAO constitution.

For example:

And so, if you imagine a longship, and every token holder as the holder of an oar, the way that I think about it is that something like a constitution is like a coxswain (the person who sits at the front of the boat).

The fact is that among the current Safe token holders, the vast majority of normal users do not have any power in fact, their participation is just wasting time with the whales without even knowing who the whales are :face_with_hand_over_mouth:

Many people think that my speech in the forum is only for the transfer of token, and that I am a short-sighted person who only considers self benefits.

I admit, it was like this at first.

And I don’t think there’s anything wrong with that, I’m not some saint and neither are you.

But when the SEP#2 vote was not passed, I started to think about the significance of my spending so much time participating in forum governance, my comments were actually on the side of the majority of users (as evidenced by the number of likes I got) , but our opinions are virtually worthless.

In fact, in SEP#2, the voting is only decided by about 5 people. I can say unceremoniously that the participation of the remaining 1800 people is meaningless, because they can’t change anything by voting or not voting, even if these 1800 people choose the same voting option (which is what it is, about 90% of people voted yes), nothing will change.

In SEP#2, two voting addresses (at present, it may be held by the same person) hold 12M voting rights, which is almost more than the sum of the current voting rights of all users.

Yet we don’t even know who those two addresses are, and we didn’t know the reason of their vote. NO transparency al all, So what is the point of other DAO members participating?

The team should at least clarify these whale addresses belong to whom, otherwise there is no need for the next vote, just an internal meeting of 5 whales can decide the proposal, please don’t let the remaining more than 1,800 DAO members participate in this meaningless vote.

To be honest, I don’t expect tokens to be transferable right now, especially in this market situation.

I’m just trying to figure out, besides the whales, what’s the point of the participation of the remaining DAO members? Is this opaque minority governance a waste of the majority’s time?

Decentralization is not just a slogan


Leaving fees open ended seems to make sense. There could be advanced features in the future, or services offered to businesses and institutions that would require fees to fund the services offered.

Dispute resolution
Leaving the dispute resolution more open-ended could be useful


…such as using Kleros Court or a similar solution…

That’s fine and all but be it as it may, the constitution is not somewhere to define fees when there are none. Use of different words or additional points may be utilized. This goes back to my comment that the principles are generalized and easy to misinterpret.

Words in bold should be revised as they open up general interpretations and/or is embellishing.

This constitution is very good, but in addition to the constitution, I think we need to formulate corresponding detailed roadmaps for safedao and safe protocol in the future, which will help dao members understand clearly what our common goal is.

1 Like

Generally supportive of the constitution and at a high-level agree with the content. A few thoughts, many of which were inspired by the replies of others:

  • I agree with many others on lowering quorum given the difficulties many DAOs have had due to low participation. I’d be open to equivalent to regular quorum or 1.5x regular quorum. 20M feels high, and we can always adjust later if the community feels as if there are too many low-importance proposals to amend the constitution.
  • Strongly disagree on adding examples to goals and principles. Examples can fall out of date quickly and can be used too rigorously for pattern matching where we want flexibility. Yes, there is a risk of stretching goals and principles in arguments, but I think this is fairly easy to recognize. In my opinion, there is greater risk that folks will reference the examples to limit changes in the future that we had not considered at this point.
  • Clarification on language is something that I think is important here. This goes beyond just the constitution, but this feels like a good place to discuss it as we are defining terms. In this draft, both “Safe” and “Safe Protocol” are used. Are these meant to be used interchangeably? I’d like to strongly advocate not using just “Safe” without a suffix (and rebranding anything that just uses “Safe”). It has been extremely confusing in conversation and I have defaulted back to saying “Gnosis Safe” to ensure folks understand what I’m referencing.
  • I agree with the spirit of the third goal (Tokenize Value), however, I’m not sure if the implementation of it makes sense for as staying a document as a constitution. Specifically, I don’t think the SAFE token should be mentioned at all in the constitution. DAO governance processes are still extremely early and it’s unclear if ERC20 token-based governance will be the end state.
  • Agree on removing fee language from first principle, both because of the implications around fees and because the term is not broad enough. “Fees” should be replaced with “Changes” as there are non-fee-based ways to compromise Safe as a public good.