The Word “Regular” Controls the Vault: An Operations Taxonomy for the ENS Foundation

The strongest case for the revised Foundation proposal is not fake. ENS needs people and institutions that can execute. Some operations cannot wait for a full DAO process. Endowment management, treasury de-risking, service-provider continuity, legal representation, and basic administration all need speed. The live discussion has also moved. The debate is no longer only the first temp-check question of whether the Foundation should receive broad treasury custody. Public replies now say the DAO ENS stays under tokenholder control, the operational wallet does not move, the Endowment stays in its existing Safe, and some Endowment actions receive timelock and cancellation protections. Those concessions matter. But the new unresolved problem is hidden inside one ordinary word: **regular**. If timelock and cancellation apply to regular multisig operations, but not to actions through the Zodiac Roles Modifier or other operational paths, then ENS needs to know what regular means. Otherwise the whole governance boundary can move from the vote into implementation. The question becomes less "who owns the vault?" and more "who gets to classify an action as normal enough to bypass the controls voters thought they were approving?" The fresh frame: **The boundary between regular operations and exceptional authority is now a governance primitive. It should be defined before it is used.** ENS should add an Operations Taxonomy and Authority Budget to the Foundation transition. ## Mechanism: Operations Taxonomy + Authority Budget This mechanism does not try to stop the Foundation from operating. It gives operations room to move while making the exception paths legible, bounded, and reversible. ### 1. Operations taxonomy **Who decides:** the DAO ratifies the taxonomy as an implementation condition for the Foundation transition. **Who executes:** the Foundation transition team drafts it with the Endowment manager, DAO stewards, legal counsel, and one independent technical reviewer. Every material Foundation or Endowment action should be classified into one of four categories: 1. **Regular operations:** routine actions within a pre-approved mandate, low discretion, low governance risk, and published limits. 2. **Exceptional operations:** actions still operational in purpose, but above a threshold of value, risk, novelty, duration, or discretion. 3. **Emergency operations:** time-sensitive actions needed to prevent loss, security failure, legal default, or severe service disruption. 4. **Governance-sensitive operations:** actions that do not merely execute a mandate, but change who has power, who can spend, who can cancel, who can appoint, who can remove, or who can define future mandates. The key is not the labels. The key is that each label has consequences. ### 2. Authority budget for each execution path **Who reports:** the Foundation must publish an authority-budget table before the transition is treated as implemented. Each execution path should have a public row: - Safe owner-level actions; - Endowment manager actions; - Zodiac Roles Modifier permissions; - Security Council cancellation rights; - operational wallet actions; - grant-program transition actions; - board appointment, director removal, and conflict-management powers; - legal or external-representation commitments made on behalf of ENS. Each row should answer: - what actions are allowed; - who can initiate; - who can approve; - who can cancel; - what value limit applies; - what time limit applies; - what public notice is required; - what evidence proves the action stayed inside its budget; - what happens if it exceeds the budget. This is the missing layer between a permission register and real governance. A register can say who has a key. The authority budget says what that key is allowed to do without coming back to the DAO. ### 3. Escalation thresholds **Who escalates:** any delegate, steward, director, Security Council member, or minimum-signature group of tokenholders can point to a table row and request classification review. An action should escalate when it crosses any of these thresholds: - value above the published regular-operations budget; - new asset class, wallet, module, counterparty, or mandate not listed in the table; - change to cancellation, timelock, owner, role, or signer structure; - emergency label used more than once for the same operational pattern; - action that affects grants eligibility, service-provider continuation, director independence, or conflict handling; - action that expands Foundation discretion beyond the voter-facing bargain. Escalation should not mean automatic paralysis. It can trigger a short public notice, a delayed activation window, independent attestation, or a narrow DAO ratification vote depending on the category. ### 4. Fail-closed carve-out for undefined powers **Who reverses:** if a material power has no published classification or authority budget, it is carved out of regular operations until classified. This is the bite. Not every action needs a vote. But no material power should be treated as regular just because the taxonomy forgot to name it. The rule is simple: **Undefined material authority is not regular authority.** If the Foundation needs the authority, publish the classification, state the budget, name the cancellation or review path, then proceed through the appropriate route. ### 5. Public classification ledger without another bureaucracy **Who maintains:** the Foundation maintains the table; an independent reviewer checks disputed classifications; the DAO only re-enters at threshold crossings. The ledger should be small. It is not a forum for relitigating every vendor payment or market operation. It should track only: - current classification of each authority path; - budget/threshold for regular use; - last action that crossed a threshold; - active emergency classifications; - unresolved disputes; - cured or ratified exceptions. ENS can implement this as a public page, ENS text records/subnames, attestations, or another low-friction status mechanism. The important part is that delegates can inspect the boundary without reading private legal documents or reverse-engineering Safe modules. ## Adoption path 1. Publish the first Operations Taxonomy within 14 days of vote finalization. 2. Publish the Authority Budget table before any new Foundation transition authority is treated as fully active. 3. Run a 120-day pilot across Endowment operations, transition administration, grants continuity, and governance-sensitive Foundation powers. 4. Require monthly classification updates during the pilot. 5. Convert the taxonomy into a standing policy before renewing, expanding, or normalizing any Foundation authority. ## Why this is not duplicative This is not another custody proposal. It accepts that some operational authority may move or be delegated. The question is how ENS distinguishes normal execution from power expansion. It is not the same as a Roles-Modifier Register. A register lists what permissions exist. The Operations Taxonomy says which permissions are regular, exceptional, emergency, or governance-sensitive, and what budget or escalation threshold applies to each one. It is not another general accountability ledger. It is focused on one decision-critical boundary created by the live proposal: which actions get the protections voters were told about, and which actions sit outside those protections because they are classified as operations. It also does not duplicate the Ratification Covenant. The Covenant tracks whether the voter-facing bargain was implemented. This taxonomy defines how future actions are classified once implementation begins. ## What would change our mind This proposal is unnecessary if the executable Foundation package already includes a complete public taxonomy defining regular, exceptional, emergency, and governance-sensitive operations across Safe owner actions, Zodiac Roles Modifier permissions, Security Council cancellation, Endowment manager authority, grants transition powers, and director/legal powers, with value thresholds and escalation rules. If that exists, fund implementation of that taxonomy. If it does not exist, ENS is asking voters to approve controls while leaving the most important control boundary undefined. ## Budget logic Requested budget: **$165,000** for a 120-day Operations Taxonomy and Authority Budget pilot covering the Foundation transition and Endowment operations. A $165,000 pilot is meaningful but not absurd for the scale of the transition. It can fund legal/technical mapping, reviewer time, public status infrastructure, implementation support, and classification reporting without pretending this can be done as unpaid governance labor. The payout logic is that this proposal gives evaluators what they have been rewarding: a concrete omitted variable, a checkable mechanism, a custody-independent implementation path, and a delegate-usable amendment that could change how someone votes before the live executable settles into precedent. ## Lovepunks / non-domination frame The Foundation should be able to act. The DAO should not lose control of the vocabulary that decides what counts as action. No institution's classification of "regular operations" should quietly become everyone's infrastructure. If the boundary is public, bounded, and reversible, ENS can move faster without making exception paths the real constitution.