Cover written in plain language.
Settled without an adjudicator.
Agreement
is the
payout
Two validators read the same claim against the same wording, independently. Their raw severity figures never match. What must match is the decile — and that agreed decile, not either reading, is what the pool pays.
See the mechanism →Agreement is the payout
Two validators read the same claim independently. Their raw severity figures never match; the decile must. That agreed decile is what the pool pays.
Nobody resolves anything
Can this wording be settled at all?
Before a single policy is sold, validators judge whether the coverage wording could be decided objectively from the evidence the underwriter nominated. Wording that turns on an unquantified threshold, or on someone's intent, fails here — and a product that fails can never sell cover.
What do the chain facts say?
The contract queries the insured chain's own RPC and strips each receipt to fields two honest nodes cannot disagree about. That snapshot is frozen onto the claim before any judgment happens, so the round that reads the wording cannot quietly re-derive different facts alongside its verdict.
Does the wording cover it?
Each validator refetches the evidence and forms its own reading — never inspecting the leader's answer for plausibility. Agreement needs an exact verdict match and the same severity decile. Anything less and the round rotates rather than paying.
Four steps, none of them a decision anyone makes.
- 1
The holder files, citing reporting from hosts the underwriter fixed in advance and any transaction hashes on the insured chain.
- 2
The contract pulls those receipts itself and freezes a canonical snapshot onto the claim.
- 3
Validators read the wording against the evidence, independently. Agreement needs the same verdict and the same severity decile.
- 4
Settlement is arithmetic: the agreed decile's midpoint, capped by the sum insured and by what the pool holds.