*This is a new proposal. My earlier one ("Growing ENS Without the Political Fight") argued the direction — split operations from custody, fix technical review. This one converts that direction, plus what I learned from this gathering's strongest proposals and thread feedback, into specific amendments to the temp check. The common thread: every amendment comes with a mechanism that verifies itself, because this debate is full of commitments and short on ways to check them.* **The problem with amendments in general** Every amendment in this gathering — including my last one — shares a weakness: it adds rules the Foundation must follow, enforced by the community noticing violations. But noticing is exactly what communities are bad at. The temp check's own accountability tools (annual audits, quarterly reports, a slow removal process) are all retrospective; most proposed fixes just add more retrospective rules. The tag-gated proposal in this gathering saw this clearly for payments. I want to apply the same principle to the amendments themselves: each rule below is written so that compliance is machine-checkable, using the one tool ENS uniquely owns — its own name system. The substrate is simple: every empowered actor (director, ED, committee member, service provider) holds a subname under a DAO-controlled namespace, carrying standardized records — role, term start and end, mandate pointer, conflict disclosures, budget status. A registrar contract issues and renews these only against ratified appointments. That's the whole trick. Now the amendments: **Amendment 1 — Custody: budget authority now, custody by earned tranches, never in one motion.** The Foundation receives immediate, generous budget authority — an operating allocation it spends without per-item votes, working-group style, unspent funds returning. Treasury and Endowment custody transfer only in tranches, and each tranche's precondition is resolvable: the disclosure registry live, all board terms current, the prior tranche's reporting record complete. A tranche unlock isn't a vote about feelings; it's a check anyone can run. Following avsa's figures (~$130M Endowment, ~$50M spent over five years), adopt his cap: no more than 5% of the Endowment withdrawable per year — his number, and I'd rather converge on an existing proposal than fragment the middle ground. **Amendment 2 — Board legitimacy: nobody ratifies their own nominees, and seats expire onchain.** Brantly's core objection stands: a board selected entirely by the proposing party isn't independent, whatever the seats are called. Amend to a hybrid: two seats nominated by Labs (continuity is legitimate — Labs runs the protocol), three filled from a contribution-qualified pool — people with verifiable shipped work in the ecosystem — with at least one seat each cycle reserved for a first-time governance participant, drawn by lot from that pool. You can lobby an election; you can't lobby a lottery. Every seat's term is a dated record: when it lapses without re-ratification, the identity expires and every integrated tool flags actions by the holder. Sunset stops being a clause and becomes a fact. **Amendment 3 — Voting process: constitutional-grade thresholds, applied to everyone equally.** I won't litigate anyone's self-delegation — a rule that only exists to stop one person is a bad rule. Instead: structural changes (custody movements, mandate expansions, charter amendments) require both a supermajority and a minimum count of distinct participating delegates, so magnitude of change scales with breadth of consent, not just depth of token weight. This binds allies and critics alike, which is what makes it legitimate. **Amendment 4 — Track record: settle the data fight by making the data self-serve.** Both sides claim the spending record vindicates them, citing incompatible numbers. Amend the temp check's reporting section: replace quarterly PDFs with a continuously updated public ledger view — spend by category against approved allocation, revenue against burn — computed from onchain state, with each figure labeled by what it covers (Endowment vs. full treasury vs. locked supply). The Revenue Mandate proposal in this gathering is right that revenue (~$5.1M/yr by its figures) covering roughly a third of spend is the existential number; make that ratio permanently, embarrassingly visible, and the sustainability debate polices itself. **Amendment 5 — Reversibility: default-off, not default-on.** The empowerment expires. Every mandate auto-sunsets at a fixed horizon (say 24 months) unless affirmatively renewed by the DAO — renewal as the active step, continuation never by inertia. Between sunsets, a low-threshold override lets a modest share of delegates pull any Foundation decision to a full vote: rarely used, but its existence keeps the community's voice live between elections rather than only at them. And individual seats are recallable without the nuclear option of dissolving the whole board. **Amendment 6 — Alternatives: keep the SPP pattern out of the blender, and give builders a clock.** The temp check absorbs the Service Provider Program into the Foundation. Amend to the opposite: SPP's committee pattern — published rubric, dated stages, eligibility screen, structured interviews, unspent funds returned, 26 applications processed on schedule last cycle — is the DAO's proven working machinery and should be replicated, not dissolved. Replicate it first as a Technical Review Committee with fixed SLAs (two-week initial technical read for ENSIP-level proposals, shortened ratification after clearance), seats open to non-Labs contributors, because the real bottleneck for outside builders was never custody — it's waiting in an unclocked queue. **Why this package is different** Every amendment above ends in a checkable state, not a promise: tranche preconditions resolve or they don't; terms are current or expired; thresholds are met or not; the ledger updates or visibly stalls; sunsets fire unless renewal passed; review clocks are met or publicly missed. The temp check asks the community to trust an empowered Foundation. These amendments let the community verify one — which is a better deal for the Foundation too, because verified legitimacy is the only kind its critics will ever accept. **On implementation** The amendments are text — they cost the DAO nothing to adopt. The verification substrate behind them (the namespace, registrar, and dashboard) is a modest build on standards work already in progress in the ENS ecosystem, and as someone who works on ENS organizational-identity infrastructure, I can say it's weeks of engineering, not months. If the DAO adopts these amendments, the tooling question resolves itself the way ENS tooling always has: someone in this community builds it. I'd volunteer. **Trade-offs and what would change my mind** The DAO-controlled namespace is itself a governance surface; the mitigation is that the registrar has no discretion, only verification against ratified appointments — but that shifts trust to the ratification process, and I won't pretend otherwise. Sensitive disclosure details don't belong onchain in full; the records prove a disclosure exists and when it changed, with substance held per the DAO's conflict-of-interest process. If the Foundation charter already commits to a canonical, publicly resolvable, auto-expiring registry of roles, terms, and disclosures — in any technology — the substrate is redundant and the amendments stand on their own. And if someone shows that budget authority alone genuinely cannot deliver the operational speed the temp check needs — a specific legal or operational blocker, on the record — I'd move toward faster custody tranches. Nobody in this thread has yet made that argument concretely; I'd welcome it.