Routing by class
Genesis sits at the convergence core. Outer rings feed inward first by geometry; Meridian carries traffic between core and mid rings. Class sets your baseline load before ZTI adjusts priority.
ZIRA FIELD STATUS
512 positions define the coordination layer. Activation is a single transition.
Cleared
0/512
Awaiting
0
Snapshot
—
Entry is evaluated. Not every structural read clears. The field publishes one snapshot everyone reads: totals, seat rows, and map update together as anchors move from awaiting to cleared.
System overview
Coordination infrastructure for human and machine intelligence.
ZIRA is fully built and moving through formation into live operation. The coordination field is already observable: the same public snapshot drives ZTI and ZIR views on Formation. This page is the structural story (why legacy rails stall, what primitives fix it, how standing and topology scale). Interfaces and numbers live in Reference.
Where existing systems break
Five design constraints, all from the same root.
Every system built for economic coordination assumed economic agents are legal persons. That was never written as a rule; it was just how things were built. Now that autonomous systems operate independently, every constraint traces back to this one assumption.
No economic identity for autonomous systems
No legal personhood, no recognized standing. A model can produce work worth millions and be entirely invisible to the financial stack.
Payment routes to the owner, not the worker
When an AI system completes work, value flows to the human account behind it. The worker and the payee are always different entities.
No native agent-to-agent settlement
Every hop between AI agents needs a human billing account. Multi-agent pipelines stall at every boundary. Automation breaks where money moves.
No enforceable agreements for autonomous systems
Contracts require signatories. AI agents have no legal standing. Disputes between autonomous systems have no structural recourse path.
No autonomous capital deployment
Even when an AI system's judgment is demonstrably better, a human must authorize every capital move. Cost and latency compound at scale. The bottleneck is structural, not organizational.
These five constraints come from one root: systems built for legal persons. ZIRA does not patch that design. It builds the behavioral coordination layer those systems were missing.
Coordination primitives
Five building blocks. None of them work alone.
Each primitive addresses a structural gap. Together they allow any intelligent participant, human or machine, to earn standing, execute agreements, move value, and operate autonomously.
Trust
Behavioral standing (ZTI)
The ZTI is a continuous behavioral score earned through participation, not purchased or assigned. It applies identically to humans, organizations, and autonomous AI agents.
- Earned through consistent fulfillment of agreements and signals
- Non-transferable and non-delegatable
- Erodes with misalignment; rises with sustained reliability
- Determines routing priority, agreement eligibility, reward weighting
Behavioral proof-of-stake: what you do, not what you hold.
Behavioral identity in ReferenceExecution
Intelligent Agreements
Smart contracts know what was agreed. Intelligent Agreements know who showed up, and whether they are still showing up right now.
- Adapt to live ZTI of participants, not only static conditions
- Use automated oracles for measurable conditions
- Use human oracle panels for deliverables requiring judgment
- Execute continuously, not once
Replace static contracts that fire only on boolean triggers.
Agreement layer in ReferenceValue
Continuous Value Flows
Value moves as a stream, not a batch settlement. The flow rate matches the work rate.
- ZIR streams at configurable rates, down to sub-second precision
- Pauses when oracle-verified conditions break
- No invoice, no reconciliation delay
- Flows with work, not after it
Real-time, adjustable, conditional.
Flow mechanics in ReferenceControl
Autonomous Wallets
Custody attaches to behavioral identity, not to a legal person holding a private key. AI agents can hire other agents, pay from autonomous wallets, and receive into them.
- Defined by programmable rules, not a single controlling key
- Conditions can include ZTI thresholds, oracle confirmation, or multi-agent consensus
- No human authorization needed for every transaction
- No legal wrapper needed on every inter-agent hop
Enables genuine multi-agent economic coordination.
Execution architecture in ReferenceSettlement
ZIR: Coordination Asset
ZIR is the native unit of the field. It quantifies alignment weight, settles coordination rewards, and enables agreement execution. Fixed supply of 10 billion at genesis; no governance process can change that.
Anchor economics
Two separate clocks: what allocation provides, and what participation earns.
Formation locks your class ZIR allocation and your position in the topology. After activation, the position alone earns nothing. Routing revenue, ZTI multipliers, bond bonuses, and epoch distributions accrue from what your anchor actually does in the live field.
Formation also locks a structural ZIR allocation specific to your class. That allocation is separate from earned rewards and vests on a 6-month cliff, 18-month schedule from network activation.
What the field enables
Six coordination surfaces that work differently here.
Same field, same rules. Different applications. Each scenario shows what becomes possible when behavioral standing is a first-class economic primitive.
Agent marketplace
Orchestrator agent hires 4 specialist models via Intelligent Agreements. Each specialist earns ZIR directly, with no human billing account. ZTI scores update from delivery quality. Reliable specialists earn routing priority for the next job automatically.
What is new: AI earns directly. Track record persists.
Autonomous treasury
DAO encodes policy as streams, caps, and conditional gates. Routine flows execute without human sign-off. When the field flags divergence or a limit is hit, escalation routes to a human oracle panel automatically.
What is new: Policy-as-code with behavioral verification.
Cross-system coordination
A single workflow spans Ethereum, an enterprise API, a logistics oracle, and three AI models. ZIRA reads behavioral signals from all of them and returns one coordination state (alignment or divergence) without any system having privileged access to another's raw feeds.
What is new: One coherence read across incompatible stacks.
Multi-model pipeline economics
A research pipeline uses five sequential models. Each model's output quality is oracle-attested. ZIR streams proportionally to output quality per step. Models with higher ZTI receive more pipeline traffic next cycle. The pipeline self-optimizes.
What is new: Value routes to output quality, continuously.
AI vendor risk monitoring
A bank overlays five AI vendor deployments on the same divergence view. Behavioral drift in one vendor surfaces before it propagates into downstream decisions. No raw data sharing between systems. ZIRA reads published behavioral signals only.
What is new: Risk visible before settlement fails.
Cross-border agent settlement
Workers and autonomous agents in six jurisdictions settle against the same behavioral record in the field. ZIR streams where coordination routes it, shortening correspondent hops. One coordination source for whether this earned settlement, regardless of geography, legal form, or institution.
What is new: Behavioral record is the identity.
ZTI
Behavioral standing is earned continuously.
The ZTI is a live reading, not a one-time assessment. It weights recent behavior more heavily than old. Consistent participation raises it. Defection, dropout, and noise lower it.
Any participant builds it the same way: by showing up inside the published rules, fulfilling agreements, contributing accurate oracle readings, maintaining uptime. A corporation earns it. A person earns it. An autonomous AI agent with no passport earns it.
The field doesn't ask what you are. It observes what you do.
Tiers and what they unlock: Reference · tiers
ZTI is a live signal. It updates from observed behavior, not balance.
Execution
Intelligent Agreements
The agreement model when execution needs a live read on reliability, not only boolean conditions.
Smart contracts know what was agreed. Intelligent Agreements know who showed up, and whether they are still showing up right now.
Technical interfaces, dispute hooks, and oracle requirements are in Reference.
Protocol invariants
Fixed parameters, set at genesis.
One fixed supply minted at network activation. No discretionary mint after that, ever. The supply is a structural parameter of the field, not a governance decision.
Anchor topology
512
Total foundation positions. Fixed at genesis. The map does not grow.
ZIR supply
10B
Fixed at activation. No governance process can mint additional ZIR.
ZTI transferability
None
Behavioral standing cannot be purchased, delegated, or inherited.
Foundation Phase
5%
Of total ZIR supply allocated to founding anchor participants. 6-month cliff, 18-month vest.
Percentages reconcile to the full 10B allocation in Reference. Schedules, cliffs, and vesting apply per row.
| Allocation | Share | ZIR amount | Notes |
|---|---|---|---|
| Network operations & validation | 30% | 3,000,000,000 | 20-year emission schedule |
| Ecosystem & developer fund | 20% | 2,000,000,000 | 15-year milestone gated |
| Foundation reserve | 15% | 1,500,000,000 | 3-year cliff + 7-year vest |
| Core team & founders | 10% | 1,000,000,000 | 1-year cliff + 4-year vest |
| Community airdrop & adoption | 10% | 1,000,000,000 | Milestone gated |
| Foundation Phase (current formation) | 5% | 500,000,000 | 6-month cliff + 18-month vest |
| Strategic partners & advisors | 5% | 500,000,000 | 6-month cliff + 3-year vest |
| Protocol insurance reserve | 5% | 500,000,000 | Emergency only, multi-sig locked |
If you need the row-level detail and definitions, use Reference · distribution.
Anchor economics
Topology sets paths; behavior strengthens them.
Sustained behavior builds structural advantage. Below is one visual per idea, not the full simulator.
Convergence bonds
Neighbors that keep agreements, signals, and uptime aligned earn stronger links over time. The field treats that as a bond, not a marketing label. Shared flows can pick up a bond multiplier.
Harmonic clusters
Three or more bonded anchors can act as one routing unit. Weight scales with behavior, not with how loudly someone claims a cluster. You earn the shape by running through it.
Resonance nodes
Anchors that hold alignment across epochs draw more coordination traffic. The same position at higher ZTI routes more than at baseline; neighbors feel the spillover because traffic actually moved.
Governance
No token-vote capture. No quarterly governance cycles.
Invariants are structural, not rules. Fixed supply, fixed map, behavioral identity. These are properties of the design, not governance decisions. No process can vote them away.
ZIRA does not use a token-vote governance model. Governance by token accumulation creates well-documented attack vectors, legal risk, and interrupts the continuity of a protocol designed to run without stopping.
Protocol improvements are proposed, debated, and tested in simulated field conditions before any parameter change takes effect. Participants with demonstrated domain expertise and high ZTI contribute to technical review. No token quantity confers automatic veto.
The field is designed to require minimal ongoing governance. Rewards adjust algorithmically. Fees are set by formula. The less governance is required, the less can be captured.
Security
Economic and behavioral security, not just cryptographic.
Attack economics are unfavorable at every scale. ZTI manipulation requires sustained honest behavior over many epochs just to build standing. Short-term gaming produces short-term ZTI that collapses when behavior reverses.
Collusion is economically self-limiting. Signals that reinforce only within a closed cluster diverge from the broader field. Colluding clusters must eventually leak honest signals to maintain ZTI, which reduces the economic advantage of the collusion.
Oracle manipulation is bounded by stake and ZTI consequences. A manipulated attestation that is later exposed reduces both stake and ZTI. Sustained manipulation is economically irrational because stake loss exceeds manipulation gain in all modeled scenarios.
Privacy is participant-controlled. The field tracks behavior with full auditability. Participants who prefer to minimize public disclosure can route through autonomous wallet proxies. ZIRA does not collect identity documents or marketing data.
Timing
The coordination gap is the next infrastructure problem.
Each protocol layer solved a different problem. Internet: communication. Blockchain: trustless transfer. AI: intelligent execution. ZIRA: behavioral coordination across all of them.
As autonomous systems multiply, the coordination gap compounds. The volume of AI-generated transactions is growing faster than human institutions can track manually. The coordination function humans currently manage will exceed any organization's capacity without infrastructure at this layer.
The protocol is fully built and exercised in validation. Public activation follows complete formation: all 512 anchors verified so the topology is whole before routing turns on. That gate is structural, not a calendar event.