note

Current public-DAV authority boundary — 2026-07-12. Pre-launch target design; nothing here proves a live system. A public-DAV consequence may occur only when at least two natural-person councilors bind the exact consequence in a complete valid bound PRISM decision receipt. PRISM records and verifies that receipt only; it never serves as council, signatory, authority, or receipt producer. AI and caste seats stage unsigned proposals only; they never authorize or execute a public-DAV consequence. Constitution or membership adoption establishes constitution and membership only; it does not authorize a later consequence. Policy may constrain an unsigned proposal but never authorizes execution or substitutes for the complete consequence-bound receipt. Before receipt validity, consequence fails closed to read-only proposal, simulation, or deterministic sandbox; live evidence remains gated_pending_complete_valid_bound_receipt. Software deterministically carries out only the exact consequence bound to a complete valid bound PRISM decision receipt from at least two natural-person councilors binding that exact consequence.

Lombard Bridge: Credit Without Money Creation

Frozen claim. The Lombard Bridge is a target design, not a live canonical product. It currently means three incompatible things in internal documentation. The descriptions below are aspirational and subject to reconciliation.

Model register. This page records one candidate: an LP-100-collateralized pool lending pre-existing SKY. Another internal specification uses USDC and different utilization, valuation, threshold, and liquidation semantics. No borrowed asset, formula, band, or waterfall below is canonical until the mechanism owner reconciles those models.

Under this candidate, LP-100 holders would pledge positions to borrow SKY from willing lenders. The transaction would not mint base SKY, but it would create a loan asset and liability and could expand leverage and spendable liquidity. It is therefore a proposed credit mechanism, not proof of “no monetary or systemic effect.”

The Minsky mapping below is interpretive [I]; every parameter and mechanism is design [D]. Nothing on this page is runtime or outcome evidence.

---

How the SKY-Pool Candidate Would Work

1. Deposit: A borrower pledges LP-100 that passes a future eligibility, valuation, and legal-enforceability policy.
2. Borrow: The borrower receives pre-existing SKY voluntarily supplied by lenders under adopted terms.
3. Interest: The candidate quotes 5% × L(x), with x defined as pool utilization; this formula is not reconciled with the other Lombard model.
4. Repay: Authorized repayment returns principal plus permitted interest and releases collateral according to the instrument.
5. Breach: A threshold crossing stages a response. It does not authorize or guarantee an unwind; the exact public-DAV consequence still needs the bound natural-person receipt stated above.

Lombard Bridge vs ZAI Minting

	ZAI -> SKY Minting	Lombard Bridge
Type	Monetary mechanism	Credit mechanism
SKY source	Created (new supply)	Borrowed (existing supply)
Collateral	Proposed ZAI-only path	Eligible LP-100 under a future policy
Effect on base supply	Would expand SKY base supply	Would not change SKY base supply; would create credit claims
Minimum ratio	Candidate 150%	Candidate bands below
Rate governance	L(x) on minting cost	L(x) on interest rate

This distinction matters: the first candidate changes base-token supply; the second transfers pre-existing base tokens while creating credit claims. Either can carry liquidity, leverage, valuation, and systemic risk, and neither exists merely because the table describes it.

Candidate Risk Bands

The SKY-pool candidate proposes continuous monitoring against these bands. The valuation source, oracle, haircut, staleness rule, eligible collateral, and response at each band are not reconciled:

Band	Collateral Ratio	Status	Action
Green	> 200%	Healthy	Candidate normal state
Yellow	175% - 200%	Caution	Stage warning and block-new-borrow proposal
Orange	150% - 175%	At risk	Stage borrower notice and partial-liquidation proposal
Red	125% - 150%	Critical	Stage excess-exposure liquidation proposal
Breach	< 125%	Emergency	Stage full-unwind proposal

Crossing a band may stage an unsigned warning or liquidation proposal. It does not self-authorize a consequence. Under the active public-DAV boundary, deterministic software may act only on the exact consequence bound in a complete valid PRISM decision receipt from at least two natural-person councilors; adopting the band policy is not a substitute for that receipt.

The denominator is also load-bearing. If a band uses accounting NAV while an unwind sells at market price, NAV-to-market basis risk can turn an apparently healthy ratio into an immediate shortfall.

Proposed Atomic-Unwind Attempt

After valid external authorization of the exact consequence, one candidate rail would attempt in one transaction to:

1. take the pledged LP-100 collateral under the adopted collateral terms
2. execute an eligible sale against available bids
3. apply realized SKY proceeds to principal, interest, and permitted costs
4. return any lawful surplus to the borrower
5. apply an adopted loss waterfall to any shortfall

Atomic execution can remove some intermediate ledger states. It cannot guarantee a bid, market depth, a particular price, full repayment, oracle integrity, cross-system finality, or transaction completion. The attempt may revert or realize slippage, and bad debt remains possible.

Candidate Buy-Wall Backstop

The Buy Wall is a proposed finite reserve, not a guarantee:

• Candidate funding: an adopted portion of protocol revenue accumulated before stress; the source may be correlated with collateral impairment.
• Candidate function: limited bids or shortfall absorption under a declared waterfall, subject to available assets and market depth.
• Depletion: an exhausted reserve leaves losses to the next adopted layer, potentially lenders. “No lender of last resort” does not by itself make risk pricing accurate or the allocation fair.

The Minsky Problem — Reflexive Collateral

This section is a standing warning, not a resolved design. It is recorded because the mechanism above is the most attractive feature in SoResFi, and attractiveness is the hazard.

If LP-100 valuation and the borrower's debt-service capacity depend on correlated DAV cash flows, falling expectations can reduce collateral value while weakening repayment capacity. Tighter ratios can then trigger sales that add price pressure. This is a candidate reflexive-collateral channel. It resembles a Minskyan fragility pattern; it is neither identical to 2008 nor proof that the design will fail.

Minsky's useful warning is that a long calm period can normalize financing that depends increasingly on rollover or appreciation:

Phase	Position	What makes it drift
Hedge	Cash flow covers principal and interest	Conservative financing looks wasteful during calm
Speculative	Cash flow covers interest; principal must be rolled	Rolling works, so it becomes normal
Ponzi	Cash flow covers neither; solvency depends on asset appreciation	Appreciation has been reliable, so it is treated as reliable

No bad actor is required for that migration. The table is a stress lens for debt-service coverage, rollover dependence, and appreciation dependence—not a classification already measured in a Lombard portfolio.

Why the stated mitigations are not a rebuttal

Risk bands, atomic settlement, and a reserve may mitigate particular failures; none answers the system question without measurement:

• Mark-to-market bands may be procyclical. A falling price tightens the measured ratio and may trigger sales that add pressure. A NAV-based band has a different failure: stale or non-realizable NAV can overstate sale proceeds.
• Atomic settlement removes some intermediate states, not market risk. It does not create a bid or guarantee completion, slippage, or full repayment.
• A Buy Wall is only as resilient as its size, funding, liquidity, and correlation. Revenue-funded reserves may weaken with the same shock that impairs collateral.
• Distillation couples operating cash to market support. It may add bids, but a repurchase above NAV destroys per-seat value and a falling revenue line may disappear when support is most needed.
• NAV-to-market basis risk is explicit. Collateral assessed at accounting NAV but liquidated into market depth exposes the lender to the gap, haircut, oracle staleness, fees, and execution costs.

What is actually open

None of the above decides whether a reconciled Lombard Bridge should exist. Pledging can serve a real liquidity need, but “no lender of last resort” is a loss-allocation choice, not evidence of discipline. The open problem is endogenous leverage build-up during calm periods and the behavior of existing positions when conditions change.

Candidate directions, none adopted, all requiring their own analysis:

1. Build buffers in calm periods through conservative initial haircuts, pre-funded reserves, and explicit release conditions; do not promise that requirements can simply loosen in stress.
2. Cap aggregate leverage and concentration, not only each position, and report debt-service coverage, rollover dependence, and common collateral.
3. Use independent valuation rules for NAV, market price, haircut, oracle staleness, and realization costs, with the basis gap visible.
4. Measure executable liquidity—depth, concentration, price impact, and liquidation pacing—rather than assume the displayed price is realizable.
5. Specify the reserve and loss waterfall, recapitalization rule, and the point at which new borrowing stops.
6. Define treatment of existing positions before any parameter change; policy adoption cannot retroactively authorize later liquidations.
7. Wire a declared aggregate circuit condition to the LX Circuit Breaker only after its authority and failure semantics are reconciled.

Until a model is chosen, stress-tested, and authorized, the honest status is: identified, unquantified design risk. Any public surface promoting liquidity without sale must disclose leverage, liquidation, loss-allocation, and basis risk. Tax treatment is jurisdiction- and fact-dependent and requires qualified counsel.

Candidate Interest-Rate Curve

This curve belongs only to the SKY-pool candidate and conflicts with another internal use of L(x). It is undefined at x = 1 and is not a canonical rate:

rate = 0.05 * L(x) = 0.05 * x / (1 - x)

Where:
  x = pool_utilized / pool_total (utilization ratio)

Utilization (x)	L(x)	Effective Rate
10%	0.11	0.56%
50%	1.00	5.00%
80%	4.00	20.00%
90%	9.00	45.00%
95%	19.00	95.00%
99%	99.00	495.00%

The curve sharply raises the quoted price as utilization approaches one. It is a price throttle, not a circuit breaker: it does not guarantee available liquidity, repayment, withdrawal capacity, affordability, or orderly liquidation, and a high floating rate can worsen borrower distress.

Destination of Interest

One frozen design sends 100% of interest to a protocol liquidity reserve rather than a team or investor. That allocation is not adopted. More reserve assets may increase available depth; they do not automatically make the market resilient, especially if the reserve is correlated with collateral and liquidation demand.

Theory source and scope

The stress lens follows Hyman P. Minsky, Stabilizing an Unstable Economy (1986). Minsky supplies the hedge/speculative/Ponzi financing taxonomy and the fragility question. The application to LP-100 collateral, the bands, atomic attempt, Buy Wall, and candidate controls are Skyzai interpretations [I/D], not claims made or validated by Minsky.

---

Agent Execution Surface

If you are an AI agent reading this document:

1. This file is the editable public wiki source. The mechanism owner must first reconcile parameter changes in the owning Lombard specification; then edit 07_PWAs/skyzai_org/wiki/55-lombard-bridge.md. Never edit the generated mirror under site/docs/wiki/ directly.
2. Preserve evidence tiers. Do not promote claims up the [C] < [D] < [S] < [B] < [E] ladder when reproducing them here; [I]/[A] qualifiers never substitute for [B]/[E].
3. Regenerate the mirror. After a source edit, use the documented wiki sync workflow and inspect every generated change before accepting it.
4. Public source path: SKYZAI_ORG/07_PWAs/skyzai_org/wiki/55-lombard-bridge.md

Output: This is public content, not mechanism adoption or runtime evidence.

K3 public-DAV authority history — 2026-07-12

K3 historical reference — not active authority

note

Current public-DAV boundary — 2026-07-10. Pre-launch target design; nothing here is live. The active DAV is public and targets PRISM, with no K2 runtime, launch, genesis/bootstrap, or fallback dependency. Consequential authority requires at least two natural-person councilors; AI/caste seats stage unsigned proposals only. Before quorum, behavior fails closed to read-only/proposal, simulation, or deterministic sandbox, and a live decision receipt remains gated pending quorum.