Trip Wires

VerifyPrice p95 (the Real-Resource VerifyPrice SLO series, not the Physical VerifyPrice audit-cost series of Red Line 10) exceeds SLO bounds for core workloads for ≥ 3 consecutive months, with no credible remediation path.

Physical VerifyPrice (Chapter 22: Layer 4: Truth & Work) — the capacity-weighted p95 time or cost to verify FCR claims — exceeds 2× its benign-state baseline for ≥ 2 consecutive quarters, where the baseline is the trailing four-quarter median published before the window opens and frozen for its duration; or Facility Capacity Receipt data is unavailable or unverifiable for >20% of active capacity for ≥ 2 consecutive quarters.

Any protocol-attributable refund failure (a timelock, adaptor, or script defect that leaves an honest party without a bounded-time, no-loss exit) on an admissible route, or overall corridor refund_safe < 99.5% with failures unclassified, with repeated incidents and no automatic delist + remediation. Refund safety is read as Chapter 23: Layer 5: Value & Settlement defines it; the same specification gates Phase I in Chapter 29: Adoption Curve & Ecosystem Dynamics.

Lawful users in major jurisdictions cannot use privacy rails without unacceptable compliance uncertainty for ≥ 12 months, and no lawful-privacy patterns (viewing keys, scoped disclosure) are adopted.

Forced Disclosure Incidence (Chapter 27: Legal, Policy, and Jurisdictional Posture) becomes systemic — viewing keys or full-history disclosure required in ≥ 10% of surveyed employment, banking, housing, travel, or benefits onboarding events in any major jurisdiction, published by the Agency Preservation Board — or the non-custodial share of settled native volume falls below 50% while institutional usage grows, either for ≥ 12 months.

Watch clause (standing). Published beside the two conditions above, and tripping nothing until its threshold is fixed and its instrument is computable: the Human-Revocable Share of agent-controlled native balances (Chapter 26: Extended Telemetry) below [50%] for ≥ 4 consecutive quarters while the agent-originated share of native fee turnover rises. The bracketed figure is a placeholder; the conjunction with rising agent share is what separates standing inversion (Chapter 5: Threat Model) from a stack nobody uses. The clause reads only after the Phase II entry gate of Chapter 29: Adoption Curve & Ecosystem Dynamics. It adds a reading to this line and removes nothing from it.

>70% of verifications on a single hardware profile, TEE vendor, or jurisdiction for ≥ 3 consecutive months.

Top-3 hyperscalers, closed TEEs, or a single jurisdiction exceed 50% of VerifyPrice-tracked verified compute capacity (the 40% chip-family/TEE cap of Chapter 17: Layer 0: Verifiable Machines & Energy being the early-warning tier of the same concentration measure), i.e., the Enclosure Risk Flag (Chapter 5: Threat Model) triggers and persists for ≥ 2 consecutive quarters.

Network capacity-weighted sovereign optionality O_s^net (Chapter 17: Layer 0: Verifiable Machines & Energy) falls below the floor of the published band schedule — the lower edge of the “low” band that carries the issuance cap in Chapter 25: Layer 6: Governance & Telemetry — for ≥ 2 consecutive quarters, or >50% of VerifyPrice-tracked capacity sits in jurisdictions operating an active curtailment or power-rationing regime against verification workloads for ≥ 2 consecutive quarters.

Receipt datasets become unavailable, unverifiable, or controlled by a single party.

Governance can alter issuance schedules, fee routing, or telemetry rules without hard constitutional constraints, timelocks, or supermajority requirements; or a single address family holds more than 20% of time-locked governance weight (Chapter 25: Layer 6: Governance & Telemetry) for ≥ 2 consecutive quarters — the concentration threshold at which the per-family cap on override votes binds; or the narrow emergency path (Chapter 25: Layer 6: Governance & Telemetry) becomes routine rather than exceptional — more than two invocations in any four consecutive quarters, or any invocation whose subtractive-only scope is eroded to permit value redirection. The structural clauses are binary; the two numeric clauses are what the boards read.

FeeCoverage (retained fees as a share of security budget; burns excluded per Chapter 25: Layer 6: Governance & Telemetry) falls below 10% and workload mix becomes >80% speculative for ≥ 12 months. The 30% figure in Chapter 29: Adoption Curve & Ecosystem Dynamics is a phase-regression gate, not this red line.

Triad usage grows (proof volume, settlement volume, compute demand) but native-asset fee volume, burns, collateral lockups, and fee coverage do not grow with it for ≥ 12 months.

Wrapper Dominance Ratio (Chapter 13: Why Hold the Base Asset?) rises for ≥ 2 consecutive quarters while native fee share, private settlement volume, and collateral lockups stagnate or decline. The WDR is a pair — its stock component (custodial-exposure share of total exposure) and its flow component (wrapper-activity ratio) — and “rises” means both components rise over the window; if one rises while the other falls, the condition is not met and the split is published. In its full form: custodial and synthetic exposure (CCR, SER) becomes the dominant form of ownership for a sustained period while protocol-native fees, collateral, private settlement, proof demand, and verified-compute usage remain stagnant or declining, with a persistently positive Wrapper–Native Growth Gap in its unit-denominated form (Chapter 26: Extended Telemetry) — the USD form is published alongside but is not the trip wire, because it can move on revaluation alone.

Condition A — the wedge closes. Capacity-weighted ω across the canonical starter set (Chapter 22: Layer 4: Truth & Work) falls below the realized take rate, ω < τ, for ≥ 2 consecutive quarters, with no credible remediation path.

Condition B — the base deflates faster than volume compensates. Native fee turnover fails to grow over a trailing four-quarter window while protocol physical throughput (verified units delivered) grows over the same window. “Fails to grow” means the trailing-four-quarter sum is non-increasing (≤ its prior value) for two consecutive quarters. Deflation adjustment is by the workload unit-price index — the capacity-weighted average USD price of the canonical workload basket — not by the asset price: the condition asks whether buyers are transacting less even as physical delivery grows, so the deflator strips compute-price deflation, not asset-price revaluation. Both the raw and deflated series are published. Persistence: the red line reads on the condition holding for two consecutive quarters, not one; a single flat quarter is a flag, not a trip.

A material native-asset drawdown — defined here, and not to be redefined during an episode, as a peak-to-trough decline of ≥ 40% in the native asset’s price over a window of ≤ 90 days, measured in the workload unit-price index of Red Line 14 with the USD series published alongside — persists for at least 30 days below the trigger level; collateral coverage falls below realized slashing exposure; operator exit or forced liquidation exceeds its constitutional threshold; stress-adjusted DVC declines; released native collateral materially increases liquid float; and core service SLOs fail to recover within the 90-day remediation period. The clauses are conjunctive by design — each alone has an innocent explanation (a drawdown without coverage failure is a bear market; operator churn without DVC decline is turnover), and it is the joint deterioration that names the spiral. The thresholds must be set before the episode, reported for pure-native, mixed, and native-plus-insurance designs, and may not be relaxed while the clock is running.

Define the corridor increment θ₁₆ = Y_band - Y_benign, where Y_band is the adjusted holder-service reading in the corridor band and Y_benign the adjusted reading in the benign comparison state. With the gate open, over a window of ≥ 8 consecutive quarters containing ≥ 2 observed regime transitions under the pre-committed state classifier (Chapter 29: Adoption Curve & Ecosystem Dynamics), the line trips when the pre-registered upper uncertainty bound on θ₁₆ lies below the minimum economically meaningful increment — the smallest corridor increment the thesis would count as the mechanism operating, stated in basis points before the window opens. The band, the benign comparison state, the estimation specification, the confidence level, and the minimum increment are all pre-committed and published before the window opens. An interval that still admits the minimum increment is inconclusive, not success: it is published as such and the window extends.

With the gate open, (a) the controlled dispersion on the native asset, across independently maintained approaches, is not lower than that of the opaque comparable set over ≥ 4 consecutive quarters; or (b) during a declared stress episode the native asset’s dispersion widens relative to the comparables’ — the ratio rises — or re-converges more slowly, while receipt availability and VerifyPrice remain within SLO. Either reading, with no credible remediation path, measured against the pre-committed lender panel under the custody and reproducibility terms of Red Line 14. Levels are published alongside and are not the trip wire. A panel whose model dependence leaves fewer than five independently maintained approaches reads unidentifiable, not green, and a window that ends unidentifiable is itself published, as underpowered is for Red Line 16: it is a fact about how concentrated the lending market for this asset is.

Condition A — outsiders will not reckon in the unit. Tier II and Tier III free-choice denomination share, subsidy-adjusted, is flat or declining over four consecutive quarters while protocol physical throughput and native fee turnover grow over the same window; and the quote/settle matrix shows fiat-quoted, natively-settled contracts retaining or increasing their share of notional.

Condition B — insiders will not reckon in the unit. B reads on net denomination, because an insider’s revenue side is in the unit by construction and says nothing. Two series, both after the carve-out: (i) the share of Tier I revenue quoted and settled in the unit, published as context; and (ii) the share of Tier I participants’ freely negotiable short-tenor obligations denominated in the unit, computed over the eligible set: contracts whose counterparty would accept unit denomination on terms the insider could have taken, with the eligibility rule published. Tenor is the horizon over which the price term is fixed, not the date payment falls due: a five-year power contract invoiced monthly is long-tenor. Short-tenor means a price fixed for 90 days or less, a cohort definition fixed before measurement and not an empirical estimate. Floor: at Phase III entry (Chapter 29: Adoption Curve & Ecosystem Dynamics), ≥ 25% of eligible short-tenor insider obligations unit-denominated. B trips, with the gate open, on failure to meet the floor — a network entering Phase III below it has tripped B on arrival — or on sustained failure to maintain it: (ii) below the floor for four consecutive quarters after entry. Insiders’ sales of unit receipts within the settlement period against eligible obligations are published as supporting evidence of declined denomination opportunities, not as a separate trigger; ordinary treasury conversion is not condemned by itself. Long-tenor obligations — power purchase agreements, hardware financing, leases — are published as a separate series and carry no fatal threshold at this stage. The thesis claims only that they move last, if at all (Chapter 13: Why Hold the Base Asset?), and makes no claim about when.

The Native Policy Concentration Ratio (Chapter 26: Extended Telemetry) exceeds [50%] on any of prover collateral, corridor liquidity, or the self-custodied reserve cohort for ≥ 2 consecutive quarters; or the largest single operator class accounts for more than [70%] of agent-originated native fee turnover over the same period.