Status: pre-launch. This page describes target design; this document does not establish a deployed service.

K1-K9: The Nine Kernel Invariants

Current reading — 2026-09-27

This page preserves a predecessor financial-kernel profile. SPLIT.PHI is not the archival K3/OUT.ARCHIVE tag. Instrument supply, financial rates, immutable classifications, enforcement and settlement guarantees are not established by calling them invariants. Applicable exact-bound external authority remains separate; see typed authority. [I]

This reader boundary does not migrate interfaces, choose unresolved terms or prove deployment. See current status.

The predecessor kernel profile below is retained for comparison. Naming a financial constraint an invariant does not establish implementation or ratify its economic terms. The formal doctrine and current cooperative-stack fences remain source-owned; this explanatory repair does not amend them.

---

The Invariants

K0: No Receipt, No Reality

Target path: every state change would produce a cryptographic receipt. Unreceipted events do not exist in the protocol's reality. This is Gate A.

• Enforcement: Target path: the transaction would revert if receipt generation fails. Current proof lanes may fall back to append-only local evidence before full external anchoring is wired.
• Scope: All token transfers, governance votes, agent actions, revenue distributions, liquidations.

K1: Conservation

Target design: Total LP-100 per DAV = 100 and Total ZAI = 100, but the "forever" invariant is contradicted by the mint path in code (frozen claim until reconciliation complete).

• Enforcement: Target design: any transaction that would change the sum of all LP-100 balances would revert (frozen claim until reconciliation complete).
• Implication: Target design: there would be no inflation, no deflation, no hidden minting. Demurrage would redistribute; it would not create or destroy (frozen claim until reconciliation complete).

Private-DAV reference (single-owner approval): typed mortal authority

Three substrates are required for every consequential action: L1 mechanical (smart-contract logic), AI/caste deliberation that remains unsigned, and the applicable natural-person authority receipt.

• Private DAV reference: under applicable law, the Natural Person authorizes their own exact typed private act. No agent receives that authority.
• Public DAV: the signer(s) required by applicable law, the entity's governing instrument, or a valid competent mandate bind the exact consequence in a complete valid bound PRISM decision receipt; PRISM records and verifies only, agents stage unsigned proposals only, and deterministic execution is limited to that exact consequence.
• Enforcement: Target design: transactions lacking any of the three substrates would revert (frozen claim until reconciliation complete).

Historical fixed financial split — unresolved research, not current implementation

SPLIT.PHI: Phi-Split (legacy bare label: K3)

Revenue is split 61.8% to equity holders, 38.2% to operations. This ratio is a constitutional constant derived from the golden ratio Nash equilibrium.

• Tag: SPLIT.PHI — not bare letter K3 (2026-08-03 naming reconciliation; K3/OUT.ARCHIVE = no silent erasure).
• Enforcement: Target path: the split would be computed on-chain. No parameter would control it.
• Implication: Neither equity holders nor operators can capture more than their equilibrium share.

OUT.EXIT / K4: Grace Exit

The retracted slogan is 'leave with everything.' The retained principle is documented exit with legitimate/redeemable property, data and keys subject to actual rights, surviving obligations and required retention. Immediate NAV liquidity and cooldown terms remain unreconciled.

• Enforcement: Target design: Grace Exit function would always be callable and could not be paused or gated (frozen claim until reconciliation complete).
• Implication: This is Gate C. Target design: the door would always be open (frozen claim until reconciliation complete).

K5: Truth Gate

Target design: Transparency score would need to stay >= 90% for capital-raising operations (Faucet, Event Tenders) to function.

• Enforcement: Target path: the Faucet contract would check a Transparency Oracle before executing. Below 90%, the transaction would revert.
• Implication: This is Gate B. No transparency, no growth.

K6: Zero Extraction

eta = 0. The protocol is designed to extract zero value beyond what is needed for regeneration. Fees exist to cover operating costs. Rent does not exist.

• Enforcement: Target path: fee schedules would be published on-chain or in equivalent governed disclosures. Fee revenue would follow the phi-split. No hidden extraction channel would be allowed.
• Implication: There is no extractive party. The organism sustains itself; it does not feed a parasite.

K7: Fair Ordering

A consensus system orders accepted transactions under its specified assumptions. Ordering alone does not establish universal absence of front-running, MEV or application-layer manipulation. The actual ledger, admission and exchange contracts need separate evaluation.

K8: Atomic Settlement

Atomic execution within a specified rail prevents its defined partial state. It does not guarantee liquidity, repayment, accurate inputs or absence of liquidation losses. Cross-rail settlement requires its own contract, finality and failure handling.

Historical immutable archetype/multiplier proposal — unresolved research, not current implementation

K9: Immutable Archetype

A DAV's archetype (PPO, Physical DAV, or Software Organism) is declared at genesis and cannot be changed.

• Enforcement: The archetype is written into the DAV's genesis receipt. No function exists to modify it.
• Implication: The L4 multiplier (1.0x, 1.5x, or 1.2x) is permanent. No "pivoting" from Physical to PPO to escape higher capital requirements.

---

Invariant Properties

All nine invariants share three properties:

1. Constitutional: They cannot be amended by governance. They are above the law of the organism.
2. Zero-discretion: No human judgment is involved in their enforcement. They are mechanical.
3. Fail-closed: Violation causes revert, not degraded operation. There is no "soft violation."

Hierarchy

K0 (Receipt)     -- foundation of all others
K1 (Conservation) -- supply integrity
Private-DAV reference, single-owner approval (Mortal) -- natural-person sovereignty
SPLIT.PHI (φ-split; legacy bare K3) -- fair distribution
OUT.EXIT / K4 (Grace Exit)   -- freedom (redeemable exit)
K5 (Truth Gate)   -- accountability
OUT.ETA0 / K6 (Zero Extract) -- non-parasitism
K7 (Fair Order)   -- equal access
K8 (Atomic)       -- settlement integrity
K9 (Archetype)    -- identity permanence

K0 is the foundation. Without receipts, no other invariant can be verified. Remaining slots build upward. SPLIT.PHI is the phi-split; do not call it bare K3.

---

Historical predecessor wording — retained on 2026-09-27, not active rules

The displaced passages are preserved verbatim for lineage. This retention date is not their original authorship date.

Predecessor passage 1

Target design: any LP-100 holder would be able to exit at NAV at any time with legitimate/redeemable assets and data export — no lock-up as doctrine, no governance veto of exit (frozen claim until reconciliation complete). Legacy slogan “Exit with legitimate/redeemable property, data, and keys” is retracted maximal wording.

Predecessor passage 2

Target design: all multi-step operations (Lombard Bridge liquidation, PRISM trades, cross-DAV transfers) would complete atomically or revert entirely. No partial states.

• Enforcement: All complex operations are wrapped in atomic transaction boundaries.
• Implication: No bad debt from failed liquidations. No stuck Algorithmic TROs. No half-executed trades.

Predecessor passage 3

The kernel invariants are the constitutional laws of every DAV in the Skyzai ecosystem. They are not guidelines. They are not best practices. They are hard constraints that admit zero discretion. No governance vote, no admin key, no emergency override can violate them.

This page states the invariants in their constitutional form. Read enforcement language below as target protocol behavior unless a narrower runtime proof has explicitly closed that lane.

Foundation source. Each of K0–K9 compiles from a specific Foundation-level invariant or axiom. The K-invariants are not policy choices — they are the generative Lagrangian written into runtime. The full compile table (K# → Foundation source, with commentary and kill criteria) is a private, unpublished canon map.

Predecessor passage 4

Target design: all transactions would be ordered by Hedera's hashgraph consensus. No front-running. No MEV. No transaction reordering.

• Enforcement: The DAG consensus layer provides this natively. It is not a smart contract property -- it is a consensus property.
• Implication: Equal access. A retail participant's transaction is ordered with the same fairness as an institutional one.

The current authority boundary above remains separate.