03The attestation primitive

Two parties. Two incentives. One proof.

A Proof of Care attestation pairs a bonded human validator's signature over a physical session with a zero-knowledge proof generated on the participant's own device. Neither part is worth anything alone.

3.1 · Formal specification

Both parts present, both consistent

σsession

Physical session signature

Signed by a certified validator in the same place as the participant. It carries a commitment to the measured value — never the value itself.

+
πzk

Client-side zero-knowledge proof

Built on the participant's device, proving the committed value satisfies the clinical rule. Observers learn the rule was met, nothing more.

=
PoC

Verified attestation

The contract checks that the proof references the same commitment the validator signed, and rejects any pair that diverges.

What the session signature commits to

  • The milestone identifier
  • A timestamp
  • The validator's accreditation reference
  • A device attestation from the point-of-care peripheral, if one was used
  • A binding commitment to the measured value, held on the participant's device
Stated plainlyA validator could sign an honest session for a participant who does not actually follow their care plan. The defense against that is economic, not cryptographic — see 3.4.
Fig. 3.1Anatomy of a Proof of Care attestation.
3.2 · Taxonomy of verifiable milestones

The milestone is the unit of settlement

Each is defined by a clinical rule, an evidence standard and a funding weight. The registry is governed by clinical review, not by token vote.

Beachhead · highest volume

Biomarker screening continuity

Periodic screening within the prescribed interval — metabolic panels, blood pressure monitoring, retinal imaging for diabetic populations.

Quorum of sessions

Medication & lifestyle adherence

Continuous supply verified against dispensation records, plus structured lifestyle sessions. A ninety-day program cannot be satisfied with one visit.

Whole household

Maternal & nutritional compliance

Antenatal visits, micronutrient supplementation and child growth monitoring — involving family members beyond the primary account holder.

Pre-authorized

Point-of-care diagnostics

Triage tests at a community node, authorized in advance by a licensed professional. The protocol records completion and the result's category — never the result.

Every registry entry specifiesA maximum attestation latency, a required validator accreditation class, and a funding weight that payer vaults use to compute releases. Adding a milestone needs clinical review — not a redeployment.

3.3 · The actuarial value recycling engine

It only has to be cheaper than the admission

Preventive actions create measurable cost avoidance for a specific payer, and that payer will pay a portion of it to have the actions verified.

Model the baseline

Actuaries project a cohort's annual cost of care from claims history — ER visits, avoidable admissions, dialysis starts.

Set the targets

The payer specifies adherence or screening improvements and the avoidance they should generate.

Deposit & lock

Capital enters an Institutional Payer Vault, with LRGN locked against the program's milestones.

Release or return

Verified attestations release rewards and validator fees; unspent balances return to the payer at contract end.

20–25%modeled reduction in preventable acute admissions among participants adherent for a full program cycle

ModeledA scenario input for planning, not an observed result. It will be replaced by measured performance as program data accumulates.

  • The reward budget exists before anyone acts — no schedule can pay more than the payer committed.
  • The payer compares a verified payout with an unverified baseline cost, not the abstract value of prevention.
Fig. 3.3How payer capital moves from deposit to verified release, or back to the payer.
3.4 · Sybil resistance & game-theoretic slashing

Fresh accounts are cheap. Accredited people are not.

The adversary is a colluding validator and participant producing attestations for sessions that never happened. Four mechanisms act together.

01

Randomized audit

Any attestation may be re-examined — by an independent validator sent to the same participant, or a consented record query.

02

Stake lockup

Bonded collateral stays locked through a challenge window that lengthens with the milestone's value.

03

Quadratic penalties

Penalties scale with the square of the mis-attested value, so one large fraud costs more than many small ones.

04

Accreditation loss

Accreditation is issued by health authorities. Lose it and you lose the ability to sign at all.

Fig. 3.4The audit and slashing decision path for a challenged attestation.

The composite effect is a floor on the cost of fraud relative to its yield.