Building Accountability Into the ENS Foundation: An Accountability Score With a Real Gate

**A note on this revision** Every review of the earlier draft said the same thing in different words: the principles were right, the mechanism wasn't there. That's a fair, unanimous hit. This version keeps only the parts of the earlier argument that survive contact with an actual formula, and deletes the rest. **The mechanism — one formula, one gate** Each quarter, the Foundation publishes a single Accountability Score, computed as an equally-weighted average of three metrics already specified by other proposals in this gathering: the Tag-Gated Accountability Layer's compliance rate (share of payments and governance actions correctly tagged and verified against the registry), the Return on Governance dashboard's mandate-delivery rate, and the Community-Signal Layer's response-completion rate (share of threshold-crossing signals answered within the required window). The DAO ratifies a minimum threshold score at the same time it ratifies the Foundation empowerment package. If the published quarterly score falls below that threshold, the Foundation's next operating tranche is capped at the prior quarter's actual spend rather than any requested increase, until the score recovers above threshold for one full quarter. No vote is required to apply the cap — it is a published-fact trigger, the same enforcement pattern "Post a Bond" already uses elsewhere in this gathering. **Why one number, not three separate reports** Each of the three source metrics already exists as a separate proposal. None of them, alone, answers the question a delegate actually needs answered in thirty seconds: is the Foundation, on the whole, performing. A single composite score with a single published threshold makes that judgment legible without requiring anyone to reconcile three separate dashboards. **How this fits with what's already funded** This is a composition layer, not a new data source — it computes a number from three metrics this gathering has already separately proposed collecting, and reuses "Post a Bond"'s published-fact-trigger pattern for the consequence. If any one of the three source metrics isn't adopted, the formula reweights across whichever of the three are live, so this doesn't hard-depend on all three passing. **Trade-offs this proposal accepts** An equally-weighted average treats all three inputs as equally important, which is a simplification — a Foundation could score well on two metrics while quietly failing the third. The mitigation is that the underlying three reports stay independently published in full; the composite score is a summary judgment, not a replacement for reading the components. A capped tranche is a real operational constraint on the Foundation, not merely symbolic — that's the point, but it means this proposal has real teeth that a purely advisory accountability layer would not. **What would change our mind** Show that a composite score obscures more than it reveals compared to reading three separate reports, and we'd withdraw the composite in favor of just publishing the three inputs side by side. Show that a hard tranche cap creates legal or operational risk the Foundation can't reasonably plan around, and we'd soften the trigger to a mandatory public explanation rather than an automatic cap. **Budget** A formula and a publishing step, computed from data the source proposals already collect. Estimated build: $1,000. **Closing** Principles without a gate are a restatement. This is the gate.