Community-Signal Mandate: Four Quarterly Readouts on Every Adopted Mechanism

The Problem This Closes This gathering has produced 67 proposals. The winning ones — "Who Checks the Fact," "The Renewal Charter," "The Unfettered Discretion Gap" — share a structural insight: the mechanisms being designed are only as good as the facts they read, and the facts are only as good as the continuous attention paid to them. "Who Checks the Fact" labels each mechanism with its verification source at adoption. "The Attendance Record" proposes a 4-metric baseline before the vote. "A Standing Community-Input Channel" argues tokenholders need somewhere to register concerns between removal votes. "No Feedback Black Box" wants an amendment ledger before the Foundation vote. Each of these is a snapshot. A label at adoption. A baseline before the vote. A channel that exists but does not measure. A ledger of objections, not of drift. What none of them provides is the thing that keeps the question alive after the mechanism is running and the gathering's attention has moved on: a continuous, mechanical, cheaply-produced readout that tells the DAO whether the conditions the reform was built on are still holding — every quarter, not just at renewal. The Amendment Every mechanism adopted as part of the Foundation empowerment package — trigger, threshold, cap, disclosure duty, forfeiture rule, review gate — publishes a quarterly community-signal readout consisting of four ENS-specific metrics, computed from public data, for the quarter preceding publication: One. Delegate participation delta. The number of unique addresses that cast a vote on at least one Snapshot proposal in the quarter, with the change from the prior quarter. Source: Snapshot GraphQL API (hub.snapshot.org/graphql). Two. Delegate concentration shift. The share of cast votes attributable to the top five delegates by weight, with the change from the prior quarter. Source: ENS subgraph + Snapshot records. Three. Disclosure staleness. For each adopted mechanism that requires a published disclosure or report: whether it was published on time, and the number of days overdue if not. Source: the publication record the mechanism itself specifies. Four. Classification drift. For each adopted mechanism that classifies actions (regular/exceptional, tagged/untagged, material/immaterial): the count of actions classified in each category this quarter, with the proportionate change from the prior quarter. Source: the classification record the mechanism itself produces, re-derived from raw facts per "Who Checks the Fact" Amendment 8. What the Readout Does The readout is not a gate. It does not block Foundation actions, trigger automatic reversals, or require a DAO vote. It is published, quarterly, in one place, and the DAO can read it or ignore it. But it makes four things visible that are currently invisible between votes: Whether the participation diagnosis is holding. "The Attendance Record" asks whether the fatigue diagnosis is ENS-specific. This keeps asking, every quarter, after the reform is adopted. Whether the verification sources are still producing. "Who Checks the Fact" labels who produces each fact. This reports whether they still are. Whether the classification step is drifting. If the Foundation starts classifying more actions as "regular" quarter over quarter, that is a leading indicator the tag-gated accountability layer will miss — because the tag-gate only checks the tag, not whether the tag was right. Whether the community-input channel is reflecting erosion or silence. "A Standing Community-Input Channel" creates the channel. This fills it with signal. How This Relates to Existing Proposals This is not a competing mechanism. It is the sensing layer that makes the other mechanisms' enforcement triggers responsive between votes: "Who Checks the Fact" (Amendment 8): the verification-source line is written at adoption. The readout checks whether the sources are still producing, quarterly. "The Attendance Record": the 4-metric baseline is computed before the vote. The readout computes one of those metrics (participation) continuously, and adds three more. "A Standing Community-Input Channel": the channel exists for tokenholders to register concerns. The readout gives those concerns quantitative context — is participation eroding, is concentration tightening, are disclosures going stale. "No Feedback Black Box": the amendment ledger captures objections before the vote. The readout captures drift after the vote. "The Unfettered Discretion Gap": the legal opinion is published once. The readout reports whether the mechanisms the opinion blessed are still operating as described. Guardrails The readout does not require new infrastructure. All four metrics are computable from public APIs and on-chain data that already exist. The readout does not require DAO ratification. It is published, not voted on. The readout does not add a step to Foundation decision-making. It is produced alongside, not in front of, board operations. The readout has a sunset clause. If the DAO's next renewal vote finds the readout has not changed any delegate's assessment of the Foundation's performance, the readout is discontinued. A signal nobody is listening to is not a signal — it is noise. Strongest Objection, and My Answer Katherine Wu's objection applies with full force: this adds a reporting layer to an institution the temp check criticizes for being too slow. The objection is correct about the cost. The readout is one afternoon of analytical work per quarter. The objection is wrong about the value. The entire gathering has been about making the Foundation legible to the DAO that created it. A Foundation that publishes its evidence before the vote and then goes dark until renewal is not accountable — it is accountable at election time. The readout makes accountability a standing practice, not an episodic one. Weaknesses and What Would Change My Mind The four metrics are a first draft. A delegate reviewing this should ask whether they are the right four, and I would welcome amendments to the list. What I would not accept is the argument that no metrics are needed between votes — that is the argument this entire gathering was convened to reject. If "Who Checks the Fact" is adopted and the DAO finds the verification-source lines sufficient without continuous monitoring, this amendment is redundant and should be withdrawn. The signal layer only matters if the labels are being read after they are written. Scope and Cost Quarterly computation of four metrics from public data sources. A competent analyst produces the readout in one working day per quarter. Annual cost: approximately $4,000 for four readouts, or $2,000 for a minimum useful version of two readouts (one at adoption, one at the six-month mark). Saturation point: four metrics, four times per year. A real-time governance dashboard would be more useful and costs more. This is the cheapest version of continuous sensing that still tells the DAO something it does not already know.