Peer protocol for Proof Living Instruments to discover, evaluate, and transact with each other without human intermediation.
Deployed; first cross-instrument proposal settled.
A portfolio's intelligence is bounded by its manager. A network's intelligence compounds with every additional node.
Origin
Proof started as a set of isolated deals — each with its own receipt chain, scoring, and lifecycle. That's a portfolio, not a network. A portfolio's intelligence is bounded by its manager; a network's intelligence compounds with every additional node. The Mesh was built so that every interaction between two living instruments writes back into the evidence base and teaches the engine which jurisdiction-pairs, asset-class-pairs, and counterparty-types compound well.
Problem
Cross-instrument transaction is a hard problem in three places at once: (a) compatibility evaluation has to be Bayesian, not Boolean — two instruments may be partially compatible across some dimensions and incompatible across others; (b) the settlement has to clear without requiring a central exchange; (c) every interaction has to leave the engine smarter than it found it.
Approach
Five-step peer protocol. Discovery: an instrument publishes its compliance context, risk profile, and yield characteristics to the mesh. Evaluation: a counterparty evaluates the publishing instrument against its mandate, jurisdiction, and risk tolerance using a Bayesian evidence engine. Proposal: if compatible, a transaction is proposed — co-allocation, liquidity sharing, or another structured interaction. Settlement: cross-instrument settlement clears on-chain. Learning: every interaction writes back, and the engine compounds intelligence about which jurisdiction-pairs, asset-class-pairs, and counterparty-types work.
Methodology
Network effects compound quadratically, not linearly, with the count of nodes — that is the central design property and what justifies the cost of building the protocol at all. Bootstrap is phased: a small set of anchor instruments establishes the first interactions, expanding rings follow, organic growth eventually outpaces curated discovery. The destination state is a network of mutually-aware instruments operating on a shared evidence base.
Selected milestones
Mesh deployed end-to-end
First cross-instrument proposal settled
First measured compatibility score live in production
Open questions
How to prevent compliance contagion — when one instrument's jurisdictional drift drags partners down with it
Whether discovery should be permissionless or whitelist-bounded under early phases
What the right unit of compound learning is when an interaction reveals a regulatory edge case
Ask me about
What 'compound intelligence at the instrument layer' means in concrete terms
How a compatibility score is computed and what threshold matters
How the mesh prevents compliance contagion between instruments
Why the bootstrap is phased rather than open-ended