Tectonic Briefing · Meta-analysis

Contingency Audit — Cycle 004

Meta-analysis of contingent forecasting under Knightian uncertainty

Briefings 091–120 · 30 dated artifacts · 17 August–12 October 2026

Headline

Cycle 4 contains 113 chains across 30 dated briefings. Every chain now states confirmation and refutation criteria, but complete fields do not guarantee a discriminating test. Public-event forecasts, comparative research propositions and intermediate milestones still need different closure rules. Seven early READ labels reuse R/H for incompatible meanings. Independent coding yields 112 Orienting, 1 Hybrid and 0 Representation reads, leaving no useful mode comparison. Primary sources satisfy two stated event criteria; 111 chains remain unresolved. Seven reviews are due but unadjudicated; future, mixed and unscheduled tests are separated. Event closure does not validate a broader causal explanation. The anomaly ledger contains 998 appearances of 112 distinct questions, so repetition cannot count as renewed observation. The next cycle should align each criterion with one proposition, preserve last-review dates, register an outcome source and time window, and turn public phenomena into mechanism-discriminating research designs. The principal research yield is a set of testable questions about renewed delegation demands, scarce checking, correlated foresight, productive disanalogy and usable correction authority.

Scope and fair comparison

The nominal window is 091–120. The corpus contains 30 dated artifacts, 113 inference chains and 998 anomaly-card appearances representing 112 distinct IDs. Every appearance remains in the machine-readable ledger. No number is duplicated in this window. The source cutoff is 12 October 2026.

This is a completed coding pass over the included corpus, with a bounded set of external outcome checks. It is not an exhaustive verification of every historical factual claim. The original full chains are reproduced below as archive evidence; their OBS labels remain claims made by the originating briefing. They are not newly verified observations of this audit.

MeasureCycle 003Cycle 004Interpretation
Dated artifacts3130Cycle 3 includes both dated 062 files.
Chains155113Use chain-level denominators.
Explicit confirmation and refutation fields0 / 155113 / 113Presence improved; quality needs independent review.
Anomaly appearances155998Cycle 4 has 112 distinct threads; copies are not new anomalies.
Original criteria independently adjudicatedNot measured on this rule2Outcome-rate comparison would mix evidence standards.

Outcome classification and partial follow-up

Two chains are Held at their original event criteria; 111 remain Unresolved (protocol: Indeterminate). The audit evaluates the test the author actually stated. A document or event forecast does not acquire a new causal-identification requirement during review. Conversely, a comparative proposition cannot be confirmed by a loosely related event. A Held event criterion is not model validation.

In 102-ie-1, the stated test needs one named institution to change its restriction level or date by the end of October. The Environment Agency documents increased abstraction restrictions, so that test is Held. In 104-ie-3, the stated test needs a published named assignment with a date. The AU records an adopted report for the Executive Council on 27–28 January 2027. That dated handoff earns a narrow Held; the stronger language about a budget or costly transfer remains unsupported. Environment Agency: drought report, 25 September–1 October Environment Agency: RPS 382, abstraction outside the authorised period African Union: defence committee report and dated Council handoff

The Fed decision is an observed endpoint, while 096-ie-2 asks whether data releases moved odds more than speeches. Roman’s instrument activation advances commissioning, while 095-ie-1 asks whether separating milestones improves cross-program judgement. The IEA statement clarifies the origin of promised stocks, while 114-ie-2 requires a later cumulative physical-release account. Each can be partly informed without being declared Held.

The remaining uncertainty has several different clocks. In 096-ie-2, the September deadline has passed: a dated policy statement cannot replace the missing comparison of market odds around data and speeches. In 114-ie-1, the 25 October runoff has not occurred at the cutoff, so the win-and-margin criterion is not yet due. In 099-ie-3, the next Chinese release is available but the chain also requires South African investment and output evidence: one branch is inspectable while the cross-country sequence remains open. In 091-ie-3, a number of controlled benchmark runs is a proposed design, not a booked future event whose nonappearance constitutes failure.

Timing statusChainsInterpretation
deadline not operationalized18Readiness, a run count or a qualitative horizon does not define an expired calendar forecast.
due unadjudicated7The original near review window has passed, but the required outcome evidence has not been adjudicated.
not yet due81The final stated window is still ahead. Qualifying early events could resolve it.
mixed due and future5A near component is reviewable while another component remains future or unscheduled.
adjudicated event2The stated original event criterion is met; stronger mechanism claims remain separate.

Due reviews: 092-ie-1, 094-ie-1, 095-ie-4, 096-ie-1, 096-ie-2, 097-ie-1, 097-ie-2. The YAML records a timing basis for every chain. Due but unadjudicated identifies unfinished evidence work; it is neither a success nor a failed forecast. No date is invented for unscheduled research.

Primary checkWhat is establishedChain boundary
Federal Reserve: September FOMC statement
2026-09-16
A unanimous quarter-point increase to 3.75–4.00 percent occurred. This establishes the decision, not whether data moved market odds more than speeches.096-ie-2, 097-ie-4
ECB: September monetary policy decision
2026-09-10
The ECB raised its three key rates by 25 basis points and cited continuing inflation pressure. The decision does not identify the relative causal influence of energy, core inflation, and institutional communication on market pricing.097-ie-2
US Treasury: G20 Chair’s Statement
2026-09-01
A dated chair’s statement exists, with a footnote identifying China’s dissent on four paragraphs. Its requested actions and reporting work do not fit an inference from silence. Implementation and subsequent migration of negotiating authority remain separate tests.097-ie-3
IEA: member-government stock-release statement
2026-10-07
The IEA links approximately 100 million barrels to stocks pledged but not yet released under the March action. This supports an accounting distinction between previous commitments and delivered barrels; it does not establish cumulative physical delivery by February 2027 or a final diesel allocation.108-ie-2, 113-ie-1, 114-ie-2
IEA: G7 energy-security meeting
2026-10-02
The IEA reports the meeting and continuing diesel pressure. A policy response was discussed; realised supply and price effects require later observations.108-ie-2, 113-ie-1
NASA: Roman instrument activation and coronagraph checks
2026-09-15
Instrument activation is a later commissioning milestone. It does not by itself establish calibrated survey science, the economic value of a robot deployment, or a cross-program relation between reporting discipline and outcomes.095-ie-1, 102-ie-3
China NBS: September purchasing managers’ indices
2026-10-08
The official follow-up provides separate production, orders, employment and size measures. The Chinese branch can be inspected; a pooled mechanism joining Chinese manufacturing and South African investment is not established by that one release.099-ie-3

Seven outcome categories remain available: Held, Held (Spurious), Failed-Y, Failed-Z, Pre-empted, Inverted and Indeterminate. Held is used only for the two stated event criteria; no failure or spurious mechanism is established. A missing publication cannot be treated as rejection, and a deadline that has not passed cannot produce a timing failure.

Break-point taxonomy

There is no identified dominant world-break type this cycle. The distribution is shown in full because an empty failure denominator is itself a limit of the evidence. Assigning unresolved chains to velocity error would manufacture the dominant result the protocol expects to find.

I · Agentic novelty0
II · Competitive recursion0
III · Coupling assumption0
IV · Buffer revelation0
V · Velocity miscalibration0
VI · Categorical drift0
VII · Spurious hit0

Denominator: zero independently adjudicated failed chains; 111 unresolved and two Held event criteria outside the failure denominator. Percentages are undefined. These zeros do not show that the seven mechanisms were absent in the world. No representative failed cases can honestly be supplied.

The active probe for Type VII asked whether a visible endpoint had been reached through the proposed mechanism. The monetary decisions, Roman activation and stock-release announcement failed the evidentiary threshold for such a claim: the endpoint alone leaves the mechanism open. They were therefore withheld from both Held and Held (Spurious). This is a stricter rule than Cycle 3’s “one named release became visible” convention.

Three construction problems can already be observed without an outcome. In 091-ie-1 a public bargaining field ends in a model-comparison criterion. In 103-ie-2 an inevitable-convergence hypothesis includes persistent divergence as a release. In 111-ie-1 the stated disposition and the condition labelled CONFIRMS point in opposite publication directions. These are specification defects, not demonstrated Type I–VII failures.

Meta-principles across cycles

mp-011 · Read mode predicts break severity.

held in abeyance. No R comparison group, one boundary H case, and only two adjudicated event criteria. No promotion or causal claim.

Evidence: 101-ie-3, 118-ie-3

mp-012 · Single-vector collapse explains the model-endogenous component of velocity errors.

not testable. No observed velocity failure is established; terminal construction alone cannot establish a failed mechanism.

Evidence: 103-ie-2, 118-ie-3

mp-013 · Template-projection failures can be disciplined.

provisional not retested. No cross-year failure established in the bounded checks. Absence of detection is not an exhaustive factual audit or proof of eliminated failure.

Evidence: 097-ie-3, 114-ie-2

mp-014 · A production discipline that removes a comparison group cannot estimate its own causal advantage from the resulting corpus.

canonical methodological constraint. Recurs across Cycles 3 and 4; promoted only as an identification constraint, not as an empirical claim about O superiority.

Evidence: 091-ie-2, 101-ie-3, 118-ie-3

mp-015 · A release field needs a criterion that discriminates the same proposition; the mere presence of CONFIRMS/REFUTES is insufficient.

provisional refined. Cycle 3 identified missing criteria. Cycle 4 repairs presence but exposes proposition mismatch. Refinement remains provisional because the mechanism of saturation has not been outcome-tested.

Evidence: 091-ie-1, 103-ie-2, 111-ie-1

mp-016 · Outcome closure must match the original test: an event can close its stated criterion without validating a broader mechanism or cross-case proposition.

new provisional. NASA, the Fed and the IEA make narrower components observable while their chains require later performance or discriminating causal evidence.

Evidence: 095-ie-1, 096-ie-2, 114-ie-2

mp-001 through mp-010 retained as previously recorded, not newly corroborated. No evidence of non-applicability is inferred from missing adjudication; retirement clocks are not advanced on unobserved outcomes.

Vocabulary curation

The curation pass removes the permanent vocabulary display, source archive and carried anomaly register before counting exact-name references. Counts are an upper bound on substantive reuse, not evidence that each invocation earns its definition. All 42 historical patterns remain in the ledger; no new meta-category is promoted.

Retire Sabbath Visibility after its prior watch. It has no analytical reference in this cycle. Keep Emergent Concealment retired. Flag for retirement: Weekend Translation, Mode-Switch Disarticulation, Sovereignty Arbitrage, Sabbath Operationalization, Cartel Dissolution. These flags distinguish non-use from theoretical invalidity. Reactivation needs fresh, independently supported instances, not a return to the display.

Governance Vacuum and Commons Enclosure retain their Cycle 3 discriminants. Name the responsible institution, missing capacity and consequence for the former; name the shared resource, literal access gate and gatekeeper for the latter. The high count of Commons Enclosure is a reason to inspect discrimination, not to assume that every access problem is enclosure.

Table Verification Asymmetry. Reopen only after at least three independently verified instantiations require a mechanism that cannot be represented by META-1 coupling failure; no promotion from repeated labels.

Inspect all vocabulary counts and curation decisions
ConceptAnalytical documentsBriefingsDecision
Observation-Action Decoupling1291, 92, 95, 99, 101, 102, 103, 104, 105, 111, 116, 119retained
Narrative-Physical Decoupling691, 92, 94, 99, 101, 112retained
Akrasia at Scale1103retained
Capability Opacity891, 92, 93, 99, 101, 102, 104, 108retained
Emergent Concealment0remain retired
Instrument Autonomy791, 92, 95, 101, 102, 103, 104retained
Scope Retreat691, 92, 98, 99, 101, 114retained
Dual-Track Maximalism391, 97, 101retained
Credential Foreclosure492, 99, 101, 103retained
Verification-Mode Asymmetry1491, 92, 93, 94, 98, 99, 100, 101, 102, 103, 104, 105, 110, 117retained
Peripheral Assertion2101, 104retained
Sabbath Visibility0retire after prior watch
Weekend Translation0retirement flag
Mode-Switch Disarticulation0retirement flag
Bypass Capture492, 94, 101, 115retained
Shadow Settlement2103, 110retained
Conditional Collapse291, 98retained
Negotiation Multiplication692, 97, 99, 101, 103, 104retained
Sovereignty Arbitrage0retirement flag
Buffer Collapse791, 94, 102, 103, 104, 106, 113retained
Chokepoint Cascade1491, 92, 93, 94, 96, 97, 98, 100, 101, 102, 103, 107, 108, 116retained
Tipping Cascade491, 92, 102, 104retained
Deadline Revelation894, 96, 97, 98, 99, 102, 103, 108retained
Keystone Removal191retained
Reversibility Asymmetry592, 98, 101, 102, 103retained
Verdict Compression891, 92, 94, 101, 102, 103, 104, 114retained
Effective-Date Convergence692, 94, 96, 101, 102, 103retained
Sabbath Operationalization0retirement flag
Commons Enclosure2791, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120retained
Optionality Arbitrage791, 92, 93, 99, 100, 102, 103retained
Paradigm Defection191retained
Process as Destination1491, 92, 93, 94, 95, 97, 98, 99, 101, 102, 104, 115, 116, 118retained
Cartel Dissolution0retirement flag
Capacity Hollowing1191, 92, 93, 94, 95, 99, 101, 102, 104, 115, 119retained
Category Collapse1091, 92, 98, 99, 100, 101, 102, 104, 107, 113retained
Governance Vacuum1091, 92, 98, 99, 101, 103, 104, 110, 111, 115retained
Constructive Ambiguity493, 99, 101, 103retained
Ceasefire Acceleration199retained
Electoral Correction391, 92, 99retained
Sanctuary Discount298, 103retained
Channel Decomposition1091, 92, 97, 99, 100, 101, 102, 103, 104, 113retained
Tail Calibration Failure592, 96, 101, 102, 103retained

The canonical taxonomy has not been edited by this audit. These are concrete curation recommendations for the separately reviewed update.

Anomaly accounting

The unit that can persist, resolve or disappear is the distinct question: 112 IDs. The full 998-appearance ledger records its dated history. Five archive closures were present before external reconciliation; a closure in a daily record is retained as a documentary fact even when its world outcome has not been independently verified here.

Audit dispositionDistinct threadsMeaning
closed in archive unverified4The record closes; independent adjudication is incomplete.
silently died23The ID vanishes without an explicit closure or successor link.
resolved2The specified missing observable is independently supplied.
open not yet due83The final register retains the question; carry-forward is not reassessment.

The G20 record is the clearest repair. Treasury’s dated chair’s statement fills the missing-document observable and records dissent. The larger question about assigned duties and implementation remains live, but “no text located” cannot continue doing its work. US Treasury: G20 Chair’s Statement

A second legitimate closure concerns the choice of a diesel-policy instrument. The IEA record identifies coordinated stock action, while the successor question about product shares and actual delivery remains open. Closure should move the question to its next observable, not erase it. IEA: member-government stock-release statement IEA: G7 energy-security meeting

The late-cycle notation “not reassessed today” is an improvement in honesty. It also makes an important limit visible: repeating eighty questions is not eighty new checks. “Persistently absent” is not assigned from those copies. Earlier records that disappeared are recorded as tracking failures even when their stated fair-review date remains in the future.

Inspect the complete distinct-anomaly ledger and histories
ID and originQuestionDispositionEvidence and next review
an-091-001-expired-comprehensive-clock
2026-08-17 → 2026-08-19
2 appearances
The Comprehensive Clock Expires While a Narrower Process Continuesclosed in archive unverifiedThe archive explicitly closes this record. Its cited outcome has not been independently adjudicated here; preserve the closure as an archival fact, not a verified world outcome.
Next: Consult the explicitly closed source record.
Review: No future date in closed record.
an-091-002-strong-output-worse-inputs
2026-08-17 → 2026-08-25
4 appearances
Strong Factory Output Arrives with Worsening Input Conditionssilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: the enlarged $4 billion buyback operations from 9 September, today's new-home-sales and consumer-confidence releases, and September production surveys.
Review: 15 September 2026.
an-091-003-resumed-count-trust-gap
2026-08-17 → 2026-08-25
4 appearances
A Resumed Count Does Not Restore the Trust Consumed by Its Interruptionsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: observer-mission final reports and adjudication records.
Review: 30 September 2026.
an-091-004-predictive-not-operational
2026-08-17 → 2026-08-25
4 appearances
A Predictive Model Remains Outside the Operational Loopsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: prospective validation or an operational forecasting pilot.
Review: 30 November 2026.
an-091-005-multiagent-fixed-order
2026-08-17 → 2026-08-25
4 appearances
A Multi-Agent System Uses a Nearly Fixed Tool Ordersilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: randomized order, role ablation, content-of-exchange arms, and equal-resource controls in any published system evaluation.
Review: 31 December 2026.
an-092-006-permits-without-starts
2026-08-19 → 2026-08-25
3 appearances
More Housing Is Authorized While Less Housing Beginssilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: today's July new-home-sales release, August conversion data, cancellations, and inventories.
Review: 17 September 2026.
an-093-007-funded-without-staff
2026-08-20 → 2026-08-25
2 appearances
Statutory Programs Continue Without the People Needed to Learnsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: staffing at the function level, the task force's report due in 90 days, resumption of the advisory committee, and the tempo of outbreak response.
Review: 30 September 2026.
an-093-008-consensus-confidence-tail
2026-08-20 → 2026-08-25
2 appearances
Faster Consensus May Produce a Worse Confidence Tailsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: matched-budget designs crossing exchange content with topology, reporting calibration and tail recovery.
Review: 31 October 2026.
an-094-009-trial-inside-unmeasured-outbreak
2026-08-25 → 2026-08-25
1 appearances
A Phase 3 Verdict Is Being Asked of an Epidemic Measured to a Factor of Threesilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: enrollment rate against case acceleration, detection-coverage improvements, and community-death share.
Review: 15 December 2026.
095-g20-commitment-gap
2026-08-30 → 2026-09-01
3 appearances
The G20 meeting exists before a joint action doesresolvedThe missing-document observable is filled by Treasury’s 1 September chair’s statement. This resolves document absence, not institutional effectiveness. The ID nevertheless disappeared from the anomaly register after 1 September.
Next: any published communiqué or chair's summary; failing that, how officials frame the December Miami leaders' summit.
Review: 3 September 2026.
095-roman-science-latency
2026-08-30 → 2026-09-01
3 appearances
A successful launch still lacks a scientific outputsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: deployment and commissioning reports, then first calibrated images.
Review: 31 January 2027.
095-roboshare-repeat-gap
2026-08-30 → 2026-09-01
3 appearances
A company-reported robot order still lacks independent confirmation and repeat economicssilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: a second paid deployment, recognized revenue, utilization, customer mix.
Review: 1 October 2026.
095-participation-authority-gap
2026-08-30 → 2026-09-01
3 appearances
Recognition expands faster than decision rightssilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: the published COP17 decision text's provisions on tenure, consent, and finance access.
Review: 30 November 2026.
095-haiti-capacity-calendar
2026-08-30 → 2026-09-01
3 appearances
The replacement election date remains conditional on a substrate outside the calendarsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: electoral-roll publication, department-level access, observer deployment.
Review: 13 November 2026.
096-hormuz-retaliation-gap
2026-08-31 → 2026-09-01
2 appearances
The vowed retaliation has not arrivedclosed in archive unverifiedThe archive explicitly closes this record. Its cited outcome has not been independently adjudicated here; preserve the closure as an archival fact, not a verified world outcome.
Next: Consult the explicitly closed source record.
Review: No future date in closed record.
096-atlantic-quiet-season
2026-08-31 → 2026-09-01
2 appearances
Peak hurricane season is arriving to an empty Atlanticsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: September formation count against climatology; any rapid-intensification event near land.
Review: 30 September 2026 mid-season check; 30 November season's end.
097-hormuz-attribution-vacuum
2026-09-01 → 2026-09-01
1 appearances
The tanker strike has no authorsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: UKMTO or CENTCOM attribution, any claim of responsibility, war-risk premia, and whether a second incident follows.
Review: 8 September 2026.
098-venezuela-control-rights
2026-09-03 → 2026-09-08
2 appearances
The Venezuela equity framework lacks disclosed control rightssilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: signed agreement terms or official summaries specifying ownership and control.
Review: 1 October 2026.
098-ai-shutdown-test-scope
2026-09-03 → 2026-09-08
2 appearances
Automated shutdown work has no public coverage testsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: a public technical test tied to named failure modes.
Review: 3 December 2026.
098-bridge-financial-close
2026-09-03 → 2026-09-08
2 appearances
Project BRIDGE has commitments but no stated financial closesilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: financial-close notice, route contract, or commissioned segment.
Review: 1 December 2026.
098-overshoot-irreversibility-ledger
2026-09-03 → 2026-09-08
2 appearances
The return pathway needs an irreversibility ledgersilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: pathway tables pairing temperature decline with removal limits and irreversible-loss assumptions.
Review: 3 November 2026.
098-food-safety-system-rate
2026-09-03 → 2026-09-08
2 appearances
Three meal incidents lack a system-level comparison ratesilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: laboratory findings, vendor audits, and comparable program rates.
Review: 3 October 2026.
099-greenland-project-coupling
2026-09-08 → 2026-09-08
1 appearances
The Arctic portfolio names sectors before it shows coupling rightssilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: a project list with budgets, owners, permits, Greenlandic approval, and common governance.
Review: 8 December 2026.
099-stoke-funded-cash
2026-09-08 → 2026-09-08
1 appearances
The $1 billion initial close lacks a public milestone allocationsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: filed financing terms, funded amount, and a completed vehicle or subsystem test.
Review: 8 December 2026.
099-unesco-implementation
2026-09-08 → 2026-09-08
1 appearances
The common-good frame has no observed enforcement mechanism yetsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: adopted text and the first national implementation commitment.
Review: 12 September 2026.
099-asiatic-election-withdrawal
2026-09-08 → 2026-09-08
1 appearances
The Asiatic Society's withdrawal notice gives no accessible causal accountsilently diedThis stable ID disappears before the final issue with no explicit closure or successor-ID link in its record. This is a tracking failure only; it does not establish resolution, persistence or failure in the world.
Next: the withdrawal document or a replacement election notice.
Review: 15 September 2026.
an100-red-sea-retention
2026-09-11 → 2026-10-12
21 appearances
No commercial response yet establishes a durable chokepoint effectopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: changed insurance terms, rerouting, or naval adaptation.
Review: 31 December 2026
an100-canada-contracts
2026-09-11 → 2026-10-12
21 appearances
Canada’s initial drone contracts have not yet demonstrated production at scaleopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Disclosed volumes, milestones, or follow-on procurement.
Review: 31 March 2027
an100-rail-feedback
2026-09-11 → 2026-10-12
21 appearances
The rail-redress proposal does not yet show a preventive feedback loopopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Automatic redress, aggregated disclosure, and a corrective-action duty.
Review: 31 March 2027
an100-measurement-decisions
2026-09-11 → 2026-10-12
21 appearances
Expanded measurement has not yet demonstrated downstream decision gainsopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A documented decision with validation evidence.
Review: 31 December 2026
an100-quantum-use
2026-09-11 → 2026-10-12
21 appearances
Shared quantum access has not yet demonstrated cumulative useopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Third-party results on shared tools, or a named foundry customer.
Review: 31 March 2027
an101-brics-transfer
2026-09-14 → 2026-10-12
20 appearances
The BRICS declaration lacks an implementation transfer mapopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A named mechanism assigning resources or duties.
Review: 31 October 2026
an101-agent-handoff
2026-09-14 → 2026-10-12
20 appearances
The enterprise AI harness lacks independent recovery evidenceopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Independent evaluation of escalation, rollback, and audit reconstruction.
Review: 31 December 2026
an101-enso-precommit
2026-09-14 → 2026-10-12
20 appearances
The El Niño forecast has not yet produced visible sector triggersopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: conditional preparations tied to regional indicators.
Review: 31 December 2026
an101-nsf-portfolio
2026-09-14 → 2026-10-12
20 appearances
NSF’s portfolio promise has no visible allocation pattern yetopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Solicitations or awards that fund structurally different work.
Review: 15 November 2026
an101-osf-stewardship
2026-09-14 → 2026-10-12
20 appearances
The OSF transition lacks a durable replacement for active contextopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Migration paths that retain identifiers, histories, and permissions.
Review: 16 November 2026
an102-cee-gas-contracts
2026-09-15 → 2026-10-12
19 appearances
Central Europe’s gas terminals have seven times the capacity of its contractsopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Long-term supply agreements reported by regional buyers.
Review: 1 January 2027
an102-sweden-late-votes
2026-09-15 → 2026-09-17
2 appearances
Sweden’s late-vote channel is large enough to move a preliminary resultclosed in archive unverifiedThe archive explicitly closes this record. Its cited outcome has not been independently adjudicated here; preserve the closure as an archival fact, not a verified world outcome.
Next: final certification, tracked under the Institutions lens rather than as an anomaly.
Review: No future date in closed record.
an102-digit5-certification
2026-09-15 → 2026-10-12
19 appearances
A humanoid for work beside people references a safety standard it is not yet certified againstopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Third-party certification before general availability at the end of 2027.
Review: 30 June 2027
an102-us-fire-ratio
2026-09-15 → 2026-10-12
19 appearances
Fire acreage at 147% of average sits under a mid-scale preparedness levelopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Whether the level rises with late-season acreage.
Review: 31 October 2026
an103-second-round
2026-09-17 → 2026-10-12
18 appearances
Two central banks acted on second-round effects that neither has yet observedopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Core and services inflation and pay settlements in the October releases and the February reports.
Review: 28 February 2027
an103-water-baseline
2026-09-17 → 2026-10-12
18 appearances
Water planning still anchors to a reference period that describes a minority of yearsopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A national water authority adopting storage-based or revised-baseline triggers.
Review: 30 September 2027
an103-paper-agents
2026-09-17 → 2026-10-12
18 appearances
A paper-agent disagreed with its paper and no publisher has said what that meansopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A publisher pilot or policy on paper-agents, and a resolution of the gene discrepancy.
Review: 31 March 2027
an103-eogs-budget
2026-09-17 → 2026-10-12
18 appearances
Europe’s sovereign imaging service has two blueprints and no budget lineopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A line for the service in the 2028–2034 framework with a satellite count.
Review: 31 December 2026
an104-forum-transfer
2026-09-18 → 2026-10-12
17 appearances
Three forums in one week, and none has yet published a transferopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A communiqué or outcome document that names an office, a resource, and a date.
Review: 15 October 2026
an104-doctor-replacement-ledger
2026-09-18 → 2026-10-12
17 appearances
WHO published the age stock of doctors without the replacement flowopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Country-level vacancy and training figures alongside the age distribution.
Review: 31 March 2027
an104-semicon-yields
2026-09-18 → 2026-10-12
17 appearances
India’s inaugurated chip lines have no published yield or customeropen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Yield and customer disclosures from the operating lines.
Review: 31 March 2027
an104-measurement-funding
2026-09-18 → 2026-10-12
17 appearances
WHO named the gaps in its violence data without a funded survey calendaropen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Funded, dated survey commitments adopting the recommended definitions.
Review: 30 June 2027
an105-evaluator-authority
2026-09-21 → 2026-10-12
16 appearances
Embedded evaluation leaves its corrective route to be specifiedopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Published access and reporting terms, an escalation procedure, and a traceable correction or disagreement.
Review: 31 December 2026
an105-ai-alert-protocol
2026-09-21 → 2026-10-12
16 appearances
A proposed international alert lacks an agreed public operating procedureopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Joint acknowledgement of contact points, activation conditions, or a drill.
Review: 31 October 2026
an105-trial-participation
2026-09-21 → 2026-10-12
16 appearances
Clinical-trial opportunity still depends on the ability to participateopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A comparative trial report connecting specified support to retention and representation.
Review: 31 March 2027
an105-watershed-service
2026-09-21 → 2026-10-12
16 appearances
Watershed fire exposure needs a matched water-service recordopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Matched evidence on fire exposure, treatment burden, and service reliability for named systems.
Review: 31 March 2027
vnu-payment-20260922
2026-09-22 → 2026-10-12
15 appearances
An outcome claim needs a visible consequence for paymentopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Published verification criteria and a disbursement linked to the assessed result.
Review: 31 March 2027
aria-interaction-20260922
2026-09-22 → 2026-10-12
15 appearances
A biological improvement needs an interaction recordopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Contained results reporting benefit and relevant ecological interaction costs.
Review: 31 March 2028
ons-ai-account-20260922
2026-09-22 → 2026-10-12
15 appearances
AI visibility needs a stable accounting bridgeopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Prototype estimates reconciled to existing accounts and a documented scope.
Review: 31 March 2028
phage-comparability-20260922
2026-09-22 → 2026-10-12
15 appearances
Personalised decisions need comparable implementation recordsopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Reports from more than one site showing common documentation and explained adaptations.
Review: 31 March 2028
an107-agent-human-disclosure
2026-09-24 → 2026-10-12
14 appearances
Agent services need a record of when a person completed the taskopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Product terms, interface labels, or task-level records stating when a human completes an agent’s task.
Review: 31 March 2027
an107-tigray-mediation
2026-09-24 → 2026-10-12
14 appearances
Ethiopia’s escalation lacks a named mediation channelopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: an African Union– or IGAD-convened meeting of named delegations, or scheduled flights resuming to Mekelle.
Review: 31 October 2026
an107-outbreak-detection
2026-09-24 → 2026-10-12
14 appearances
The Ebola count needs a detection-adjusted companionopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Published estimates for the Congolese epidemic that pair confirmed cases with testing volume or adjust for access.
Review: 31 March 2027
an107-effective-protection
2026-09-24 → 2026-10-12
14 appearances
Protected-area targets need effective, not only designated, coverageopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Commitments at the Yerevan biodiversity summit that attach enforcement budgets and monitoring to designated areas.
Review: 15 November 2026
an108-agent-incident-route
2026-09-25 → 2026-10-12
13 appearances
Agent incidents need a reporting route with a deadlineopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: a named authority and time limit for AI developers to report agent incidents that affect public systems.
Review: 31 March 2027
an108-diesel-instrument
2026-09-25 → 2026-10-12
13 appearances
US diesel policy needs a named instrumentresolvedThe IEA record identifies the coordinated stock-release instrument. Delivery and product accounting remain open in an114-diesel-accounting; the closed policy-choice record is not a resolved price mechanism.
Next: Consult the explicitly closed source record.
Review: No future date in closed record.
an108-afpak-channel
2026-09-25 → 2026-10-12
13 appearances
Pakistan–Afghanistan exchanges lack a convened channelopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A dated meeting of both governments under a named third party, or a monitored pause.
Review: 31 October 2026
an108-liming-inventory
2026-09-25 → 2026-10-12
13 appearances
Liming’s carbon balance needs an accounting methodopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Inventory guidance or a crediting protocol that counts liming against acidity inputs.
Review: 31 December 2027
an109-transfer-capacity
2026-09-26 → 2026-10-12
12 appearances
Oil transfers: capacity or fewer arrivalsopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Weekly completed transfers and waiting-time distributions for the Gulf of Oman, alongside arrivals and repeated use of alternative sites.
Review: 30 November 2026
an109-agent-sponsor
2026-09-26 → 2026-10-12
12 appearances
Agent outreach: a counterparty who can accept responsibilityopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A platform or recipient institution implements verifiable human or institutional sponsorship, with documented authorized collaborations and data terms.
Review: 31 March 2027
an109-lab-state-checks
2026-09-26 → 2026-10-12
12 appearances
Laboratory interfaces: readiness and recovery across sitesopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Independent laboratories publish recurring use, rejected commands, and interrupted-run recovery with portable permission and readiness checks.
Review: 31 March 2027
an109-armenia-repeat-exporters
2026-09-26 → 2026-10-12
12 appearances
Armenian preferences: repeat shipments by new exportersopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Customs or exporter evidence distinguishes growth among incumbents from new firms making repeated EU shipments after the preferences take effect.
Review: 30 June 2027
an110-hormuz-first-step
2026-09-28 → 2026-10-12
11 appearances
Hormuz proposals: an agreed first stepopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: a mediated text that names which obligation is performed first, or pairs obligations on the same date.
Review: 31 October 2026
an110-pause-exit-test
2026-09-28 → 2026-10-12
11 appearances
Training pauses: a published exit testopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A developer publishes the safeguards and criteria for resuming training, with an external party able to check them.
Review: 31 December 2026
an110-screening-error-rates
2026-09-28 → 2026-10-12
11 appearances
Voter screening: published error ratesopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: DHS or participating states publish SAVE match and error rates, and India’s review committee publishes findings with counts of affected entries.
Review: 30 November 2026
an110-lithium-series-bridge
2026-09-28 → 2026-10-12
11 appearances
Lithium inventories: a bridge between old and new seriesopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: SMM publishes a back-cast or reconciliation between its old and new stockpile series, or major producers adopt an alternative benchmark.
Review: 31 December 2026
an111-eval-before-release
2026-09-29 → 2026-10-12
10 appearances
Frontier evaluations: results before releaseopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A government evaluator’s results published before a frontier model ships, with the developer’s response.
Review: 31 March 2027
an111-return-monitoring
2026-09-29 → 2026-10-12
10 appearances
Returns to Myanmar: independent monitoringopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: UNHCR or another neutral body gains access to returnees from Malaysia inside Myanmar, or the programme is suspended.
Review: 31 December 2026
an111-ethiopia-foreign-support
2026-09-29 → 2026-10-12
10 appearances
Ethiopia’s named backers: independent evidenceopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: independent documentation of foreign supply to either side, or a cross-border incident between Ethiopia and Sudan.
Review: 15 November 2026
an112-strike-tolls
2026-10-01 → 2026-10-12
9 appearances
Cross-border strike tolls need an independent countclosed in archive unverifiedThe archive explicitly closes this record. Its cited outcome has not been independently adjudicated here; preserve the closure as an archival fact, not a verified world outcome.
Next: Consult the explicitly closed source record.
Review: No future date in closed record.
an112-somalia-replacement
2026-10-01 → 2026-10-12
9 appearances
A drawdown from Somalia needs a named replacementopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: The African Union mission or the Somali government names forces to take over Ethiopian positions, or Ethiopia confirms its troops will stay.
Review: 31 December 2026
an112-merits-deferred
2026-10-01 → 2026-10-12
9 appearances
Standing dismissals leave the merits for a later caseopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A party with clear standing challenges article 154 of Decree 70/2023, Argentina’s Congress votes on it, or a petition follows the verdict in the Philippine impeachment trial.
Review: 31 March 2027
an113-override-protection
2026-10-02 → 2026-10-12
8 appearances
Mandated human review needs protection for the reviewer who overridesopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: a California statute, regulation or collective agreement protecting workers who override automated recommendations, or published override rates under SB 947 after it takes effect on 1 July 2027.
Review: 31 December 2027
an113-protein-provenance
2026-10-02 → 2026-10-12
8 appearances
Protein watermarks need a checkpoint that reads themopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A DNA synthesis provider, a biosecurity screening consortium or a protein database announcing that it will detect SynthIDBio or comparable marks.
Review: 30 September 2027
an113-humanoid-paid-work
2026-10-02 → 2026-10-12
8 appearances
Humanoid shipments need a count of robots doing paid workopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: IDC or vendor data that separate paid industrial and service deployments from research, education and exhibition use, with repeat orders by named customers.
Review: 31 March 2027
an113-copper-cause
2026-10-02 → 2026-10-12
8 appearances
Chile’s copper decline needs attribution by causeopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Cochilco or company reports that attribute lost output to ore grades, weather, sulphuric acid supply or labour action.
Review: 31 January 2027
an114-diesel-accounting
2026-10-05 → 2026-10-12
7 appearances
The 100 million barrels need a diesel figure and member sharesopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: The IEA or G7 governments publish the release by product and by member, and say how much of the total counts earlier commitments.
Review: 31 January 2027
an114-poll-count-gap
2026-10-05 → 2026-10-12
7 appearances
Brazil’s poll miss needs an account from the pollstersopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Datafolha or another firm publishes a review of its first-round estimate or a changed weighting method, or runoff polls can be compared with the 25 October count.
Review: 30 November 2026
an114-final-submit
2026-10-05 → 2026-10-12
7 appearances
A software step that finalises nominations needs a legal statusopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: South Africa’s Constitutional Court rules on whether clicking “submit as final” is a requirement of the Municipal Electoral Act, or the commission writes the step into regulation.
Review: 15 November 2026
an114-measles-coverage
2026-10-05 → 2026-10-12
7 appearances
Bangladesh’s measles count needs vaccination statusopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: The health directorate or the WHO publishes vaccination coverage for the affected districts, or the vaccination status of confirmed cases.
Review: 31 January 2027
an114-irkutsk-cause
2026-10-05 → 2026-10-12
7 appearances
The Irkutsk laboratory death needs a named causeopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Rospotrebnadzor or another authority publishes a laboratory result naming the pathogen, or observation of contacts ends with no further cases.
Review: 31 October 2026
an115-orbital-use
2026-10-06 → 2026-10-12
6 appearances
Orbital compute needs an outside task recordopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A named external workload with conditions, communications costs, interruption recovery and a measured benefit.
Review: 6 October 2027
an115-genomics-decision
2026-10-06 → 2026-10-12
6 appearances
Faster genomics needs the complete decision intervalopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Multi-site time from presentation to an actionable clinical decision, with preserved diagnostic quality.
Review: 31 December 2027
an115-ageing-record
2026-10-06 → 2026-10-12
6 appearances
Care policies need a longitudinal population recordopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Funded longitudinal monitoring linked to care access and functioning, with comparable country denominators.
Review: 31 December 2027
an115-forest-composition
2026-10-06 → 2026-10-12
6 appearances
An edge metric needs the transitions behind its signopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A comparative policy or restoration study testing decisions based on total edge length against decisions retaining patch histories.
Review: 31 December 2027
an116-afcra-use
2026-10-07 → 2026-10-12
5 appearances
AfCRA needs a method and a decision another actor can inspectopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A published rating method, independence procedures, a named initial rating and documented issuer or investor use.
Review: 31 March 2027 African Union: credit rating agency launch, 7 October
an116-children-care
2026-10-07 → 2026-10-12
5 appearances
The child-care count needs a stable follow-up populationopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A cohort or reproducible denominator distinguishing reunification, alternative care, continuing separation, coverage change and access to support.
Review: 31 March 2027 UNICEF: children orphaned by Ebola, 7 October
an116-probe-mechanism
2026-10-07 → 2026-10-12
5 appearances
The revised drug mechanism needs an independently tested rangeopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: An independent mechanistic comparison using an orthogonal intervention and a resistant target or equivalent rescue, with documented context and boundary.
Review: 31 December 2027 Nature Chemical Biology: the revised mechanism, 7 October
an116-signature-setup
2026-10-07 → 2026-10-12
5 appearances
Threshold signing retains a setup and continuity questionopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: An independently examined implementation with stated dealer trust, setup verification and a documented recovery or refusal route after abort.
Review: 30 September 2027
an117-cable-operating-load
2026-10-08 → 2026-10-12
4 appearances
The cable temperature result needs a matched operating trialopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: An independently examined energized comparison reporting safe load, conditions, durability and other binding constraints.
Review: 31 October 2027
an117-compressed-goal-transfer
2026-10-08 → 2026-10-12
4 appearances
The compressed-goal account needs a discriminating goal-change comparisonopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Preregistered transfer predictions, matched rival models and a new controlled comparison after a specified goal change.
Review: 30 April 2027 Nature Communications: compressed reward functions, 8 October
an117-procurement-inexperienced
2026-10-08 → 2026-10-12
4 appearances
Improved procurement entry leaves the next transition for inexperienced firms openopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Registration, bids, awards and contract completion for the same initially inexperienced firms, with treatment denominators retained.
Review: 30 April 2027
an117-permafrost-accounting
2026-10-08 → 2026-10-12
4 appearances
The permafrost trial needs released methods and an emissions accountopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Public methods and data, matched multi-site physical comparisons, and independently examined net-emissions accounting.
Review: 31 October 2027
an118-sudan-scope
2026-10-09 → 2026-10-12
3 appearances
The Sudan embargo’s scope has been deferred twiceopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A Council text that states the embargo’s geographic reach and its treatment of unmanned aircraft, or a third rollover with unchanged language.
Review: 11 December 2026 Security Council Report: Sudan sanctions vote
an118-asylum-procedure
2026-10-09 → 2026-10-12
3 appearances
South Africa’s court order has no operating procedureopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A published replacement intake procedure, or an application to court alleging non-compliance.
Review: 28 February 2027 Daily Maverick: asylum directive withdrawn
an118-enzyme-function
2026-10-09 → 2026-10-12
3 appearances
The AI-reported viral system needs an assay and a search recordopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Biochemical evidence of a defined activity, and the prompts and constraints supplied to the agents.
Review: 30 April 2027 CNN: the AI-led biological finding
an118-finland-baseline
2026-10-09 → 2026-10-12
3 appearances
The Finnish assessments must describe land that was cleared before they finishedopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Final assessment reports that state how the cleared area is treated, and any remediation or compensation condition.
Review: 31 March 2027 CNBC: Finland halts Google site work
gm-us-refinancing-pressure
2026-10-11 → 2026-10-12
2 appearances
US fiscal pressure is rising without demonstrated current dysfunctionopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: The November refunding plan, followed by effective interest costs and consistent auction observations.
Review: 2026-11-04.
gm-nonbank-funding-amplification
2026-10-11 → 2026-10-12
2 appearances
Nonbank vulnerability needs evidence of an activated funding channelopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Joint observations of repo terms, margin demands, forced sales and market depth over at least a month.
Review: 2026-11-11.
an119-open-data-verification
2026-10-11 → 2026-10-12
2 appearances
Faster discovery may transfer work into checkingopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: A comparison of completed dataset decisions with error and correction time, retaining rejected candidates.
Review: 2027-01-11.
an119-community-care-continuity
2026-10-11 → 2026-10-12
2 appearances
A discharge count can conceal where support obligations moveopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Longitudinal housing, care, autonomy and unpaid-burden measures tied to a named implementation.
Review: 2027-10-11.
an119-multilingual-address-route
2026-10-11 → 2026-10-12
2 appearances
Formal identifier validity can fail downstreamopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: End-to-end multilingual registration, message delivery and account recovery after the 2027 activities.
Review: 2027-06-01.
an120-robotic-score-transfer
2026-10-12 → 2026-10-12
1 appearances
A robotic skill score may change the behavior it measuresopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: An independent outcome evaluation before and after users receive score-based feedback, with transfer to new operators and settings.
Review: 2027-04-12.
an120-clinical-critical-cases
2026-10-12 → 2026-10-12
1 appearances
Aggregate documentation agreement leaves consequential cases unresolvedopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: An external evaluation reporting severity-specific errors and independent review of unflagged assessments.
Review: 2027-04-12.
an120-rank-correction-propagation
2026-10-12 → 2026-10-12
1 appearances
A corrected ranking may travel more slowly than its earlier claimopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Subsequent public summaries and investment-promotion materials that either incorporate or repeat the affected ranks.
Review: 2027-01-12.
an120-orbital-inspection-scope
2026-10-12 → 2026-10-12
1 appearances
A mission can learn that its initial inspection task needs revisionopen not yet dueThe final register carries the question forward. Most carried cards explicitly say not reassessed; carry-forward is not fresh evidence that the absence persists. Review against the stated observable and date.
Next: Published inspection results, target-condition surprises, approach revisions and the allocation of human intervention.
Review: 2027-06-01.

The YAML includes every dated appearance, original text, source file and later same-ID occurrences. The full occurrence ledger is retained there so the readable page does not make repeated cards appear to be independent evidence.

Theoretical implications

The cycle’s strongest theoretical implication concerns what an observation can close. Under an unstable or contested classification, a realised event can be compatible with several mechanisms and several definitions of success. A multi-path field can preserve this ambiguity, but it does not resolve it. A useful orienting practice therefore pairs a bounded disposition with an explicit rule for what remains unknown after each observable arrives. The daily corpus documents this design problem; it does not estimate the superiority of one epistemic stance.

The LLM-cognition implication is correspondingly narrow. Valid-looking field names can coexist with semantic drift, and a fluent terminal sentence can test a different proposition from the opening hypothesis. Those are observable features of generated text, not measurements of an internal cognitive mechanism. Identifying mean-trajectory regression, reflexive blindness or mechanism–outcome decoupling requires comparisons that separate these accounts from ordinary evidence gaps. Neither the absence of resolved failures nor the presence of polished counterfactual prose does that work.

For human–AI analytical ensembles, the audit suggests that periodic reframing must sometimes change the unit of evaluation itself. The move from event forecasts to research propositions makes an unchanged forecast score misleading; the move from new anomaly cards to carried registers makes unchanged counts misleading. Human judgement adds value when it notices that the instrument is measuring a different object. The six research designs below turn that observation into questions that can be tested rather than treating this corpus as model validation.

Research yield: from public phenomena to discriminating studies

These connections use the public research programme and the substantive question of the named studies. Each separates a sourced phenomenon from a new hypothesis. A geopolitical configuration with three actors is not, by itself, a connection to the Three-Body study: the relevant mechanism is renewed epistemic demand and the resulting composition of delegated work.

Which work returns to the founder after automation?

Three-Body ABM · moving tasks and delegation

Phenomenon. Autonomous systems are being developed alongside validation systems, while organisations face separate questions about authorisation, intervention and the conditions of use. The ARPA-H programme makes those stages explicit. The phenomenon is the recurrence and relocation of work around automation; a press release does not establish how often that work returns.

Mechanism and rival. Hypothesis: the founder’s actions and competitors’ outputs renew verification, stakeholder support and interpretation demands even at unchanged machine capability. Rival: tasks are fixed and a constant checking cost, a legal human-signature requirement or a binding time budget explains the residual human role.

Study question. When does machine performance change the reference against which the next piece of work is judged, causing a previously delegated demand to return to the founder?

Computational discriminant. At equal capability and matched workload, independently freeze reference movement from the focal venture, rivals and the external environment. Keep customer competition active while freezing referents. Compare turnover in the identity of delegated demands, not only the overall AI share. Separately vary the cost of checking and required human sign-off. A stable AI share with referent-timed turnover is the discriminating signature to investigate.

Empirical observable and design. Follow identifiable work episodes within ventures adopting agents. Record the original demand, performer, later reopening, reason for reopening, model version, customer feedback and rival releases. A matched or staggered rollout can test whether hand-backs track changed evaluation criteria after controlling for errors, regulation and capability. Retain cases that receive no stakeholder response as missing judgment, not failed quality.

Limits. Clinical robotics supplies a boundary case, not an entrepreneurship result. The proposed computational contrast represents specified sources of movement inside a model; it cannot demonstrate unimagined possibilities. No study outcome is claimed here.

Fresh question. Can better automation increase turnover in who performs each demand while the average division of labour appears settled?

Archive anchors: 093-ie-3, 095-ie-2, 107-ie-2, 113-ie-2, 120-ie-1. ARPA-H: autonomous stroke-intervention programme World Bank: Making AI Work in Europe and Central Asia

When do cheaper answers exhaust the institution’s ability to check?

GCM AI Agents · cheap answers, scarce checking and organisational attention

Phenomenon. The Accenture–Anthropic announcement makes an embedded evaluation function publicly visible. The audit’s related chains distinguish producing an assessment from changing a release decision. That distinction generates a concrete organisational question without treating the partnership as effective or ineffective.

Mechanism and rival. Hypothesis: inexpensive solution production increases demand on scarce checking and problem-sensing capacity; apparent completion can rise while unresolved problems accumulate. Rival: outcomes deteriorate because evaluators lack decision rights, even when checking capacity is sufficient. A third rival is harder incoming tasks.

Study question. Does pseudo-resolution arise from an attention queue, from missing corrective authority, or from their interaction?

Computational discriminant. Cross solution-arrival rate, verification service capacity and the authority to reopen or halt a decision while holding task difficulty and total effort fixed. Track unresolved problem stock, age, recurrence, accepted false solutions and diversion of attention from problem sensing. Authority-only and capacity-only interventions separate mechanisms that an aggregate completion rate would combine.

Empirical observable and design. In a controlled organisational task, vary answer availability and escalation rights separately. Record proposed, checked, accepted, reopened and durably resolved cases, with planted errors and delayed follow-up. For field work, use an intake-to-correction event log and compare equivalent cases around a change in escalation policy; self-reported evaluator activity is insufficient.

Limits. The public announcement verifies an arrangement, not actual veto power or queue failure. Synthetic organisations isolate represented processes; they do not establish prevalence in real firms.

Fresh question. Can adding evaluators reduce durable resolution by legitimising more unchecked solutions when their findings remain advisory?

Archive anchors: 100-ie-2, 105-ie-1, 111-ie-1, 119-ie-2. Accenture and Anthropic: embedded evaluation partnership

Does a shared machine insight retain value when everybody acts on it?

Glimpse ABM · AI foresight and entrepreneurial competition

Phenomenon. The Interaction Tax study reports solution convergence after agents exchange complete outputs under matched budgets. This supplies a demonstrated micro-level convergence mechanism in bounded optimisation tasks. The market chains raise a separate question about common AI-directed investment; they do not establish that common foresight caused market crowding.

Mechanism and rival. Hypothesis: correlated machine-assisted foresight directs firms toward similar opportunities, so better identification raises gross opportunity quality while competition reduces appropriable value. Rival: costly execution or a supply bottleneck, rather than convergence, explains weak returns. Information accuracy and cross-actor error correlation must therefore vary independently.

Study question. When does an epistemic improvement cease to be a private advantage because its recommendations are shared?

Computational discriminant. Cross forecast quality with signal correlation while holding marginal signal quality fixed; then vary adoption share and market expansion. Compare foreseen opportunity value, executed novelty and realised value. Ablate output imitation without removing shared information. Crowding should depend on correlated action and rival adoption; an execution bottleneck should survive when choices are decorrelated.

Empirical observable and design. A market experiment can randomise whether entrants receive a common high-quality signal or equally accurate independent signals, then observe entry, choice similarity, price and realised surplus. Analyse market-level spillovers instead of treating entrants as independent observations. A field design would require traceable recommendation exposure and demand changes, which the briefing does not supply.

Limits. A verifier-scored solution task is not a customer market. The proposed bridge is a mechanism hypothesis with an explicit missing step: similar answers must produce competing actions for convergence to dissipate value.

Fresh question. When can integrating different machine perspectives protect entrepreneurial value by preserving action diversity, even if it adds no predictive accuracy?

Archive anchors: 094-ie-2, 112-ie-4, 113-ie-4, 120-ie-2. Ann, Liu and Tan: The Interaction Tax

When should a useful representation be abandoned?

Polymathy LLM-ABM and orienting under uncertainty

Phenomenon. Molinaro and Collins’ original preprint tests a compression account of goal-dependent learning; the multi-agent study tests how communication changes diversity. Together they motivate a bounded question about the cost of efficiently retaining the wrong distinctions after a goal changes. Neither identifies an LLM’s internal representation.

Mechanism and rival. Hypothesis: integration earns its value by preserving source constraints and detecting when the receiving problem requires distinctions that a compact prior representation discarded. Rival: apparent polymathic gains come from more retrieval, more compute, persuasive language or a generic ensemble effect.

Study question. Does productive disanalogy improve the ability to revise a representation when the goal changes?

Computational discriminant. Cross breadth, domain depth and integration with a controlled goal switch. Hold tokens, tools and accepted evidence constant. Compare full-solution sharing with independent proposals and constraint-focused critique. Include an ablation that preserves factual access but removes the ability to reject an analogy. Measure selective transfer, invalid transfer and recovery, not just average fluency.

Empirical observable and design. Construct transfer tasks with matched surface similarity and independently specified shared constraints. Require a predicted boundary before feedback. Blind assessors to condition and test whether a disanalogy intervention improves held-out goal-change performance beyond a resource-matched generalist baseline.

Limits. The accessible source is the 2025 preprint; the October 2026 publisher page did not open in this audit. The task manipulates represented goal changes. It does not operationalise an unrestricted Knightian outcome space.

Fresh question. Is the most useful form of breadth the capacity to find another analogy, or the capacity to know which existing analogy to stop using?

Archive anchors: 091-ie-4, 092-ie-4, 094-ie-2, 117-ie-1. Ann, Liu and Tan: The Interaction Tax Molinaro and Collins: reward-function compression, original preprint

Does a warning become useful through authority or through better information?

Human–AI ensembles · epistemic coupling and persistent augmentation

Phenomenon. The G20 chair’s statement is an inspectable document that replaces an alleged absence, yet its publication does not establish implementation. Embedded evaluation similarly provides a visible information-producing role while leaving the effect of contested findings to be observed. In both cases, access to evidence and authority over action are separate.

Mechanism and rival. Hypothesis: improved sensing changes outcomes only when an actor has a usable correction route and accepts its costs. Rival: the information itself is inadequate, so additional authority cannot improve the decision. Coordination latency is a third explanation.

Study question. Under what configurations does the human contribution remain interpretation and intervention rather than ceremonial approval?

Computational discriminant. Cross signal quality, correction rights and delay under matched capability. Give some agents accurate dissenting evidence but no authority, and others authority with noisy evidence. Measure consequential corrections, justified overrides, false halts and downstream costs. A coupling mechanism predicts an interaction between usable evidence and authority.

Empirical observable and design. Use a preregistered decision experiment with auditable cases that sometimes require overriding a recommendation. Randomise both evidence access and the cost or protection of override. Report justified and unjustified overrides separately. In field records, trace one finding through a named decision maker to a changed action.

Limits. Frequent overrides are not automatically good, and agreement is not automatically abdication. These are proposed comparisons, not findings about any named organisation’s actual governance.

Fresh question. Can perfect transparency worsen abdication by making a decision appear supervised when the observer cannot change it?

Archive anchors: 101-ie-3, 104-ie-2, 105-ie-1, 116-ie-3, 117-ie-3, 120-ie-3. US Treasury: G20 Chair’s Statement Accenture and Anthropic: embedded evaluation partnership

When does an audit record cease to be an observation?

Classification limits and Knightian uncertainty · configurational computational theory

Phenomenon. The cycle repeatedly copied anomaly cards while later issues explicitly marked them not reassessed. It also used valid-looking READ tokens for incompatible meanings. These are directly inspectable properties of the archive, distinct from whether the underlying public events occurred.

Mechanism and rival. Hypothesis: a stable schema can preserve stale or semantically shifted beliefs by making record continuity resemble renewed evidence. Rival: the information remains accurate and only the presentation is repetitive. A third rival is a rational review cadence under long outcome lags.

Study question. How should a human–AI system allocate attention between new signal acquisition, old-claim reinspection and revision of the categories themselves?

Computational discriminant. Model finite attention divided among novelty, due-date reviews and category revision. Cross persistence of records with evidence decay and world change. Keep observation cost fixed; compare copying, fixed schedules and discrepancy-triggered review. Score detected category failures and missed changes separately from forecast endpoints.

Empirical observable and design. Randomise audit interfaces to display last-carried and last-substantively-reviewed dates separately. Seed known changes and unchanged long-horizon cases. Measure detection, false closure, review time and category repair. A shadow experiment can compare O/R/H formulations on the same cases without degrading the public briefing.

Limits. The daily corpus is selected by its own production process, uses no random assignment, and has correlated repeated threads. It can reveal process defects and generate tasks; it cannot establish the causal superiority of orienting or validate a model of Knightian uncertainty.

Fresh question. Does faster record accumulation eventually reduce a system’s capacity to notice that its categories no longer fit?

Archive anchors: 095-ie-4, 097-ie-3, 111-ie-1, 119-ie-2. US Treasury: G20 Chair’s Statement

Five recalibrations for the next 30

Give each chain one focal proposition and label its test as an event follow-up, a comparative mechanism test, or a proposed experiment.

Basis: 091-ie-1, 093-ie-1, 111-ie-1

Tradeoff: Fewer broad omnibus chains; some useful analogies move into the research section.

Verify in 121–150: Every 121–150 chain has a proposition ID, test-unit field, matched CONFIRMS/REFUTES pair and a named outcome source.

Apply the original event or comparative criterion, retain partial milestones and leave missing evidence unresolved.

Basis: 095-ie-1, 096-ie-2, 114-ie-2

Tradeoff: Fewer apparent resolutions and more delayed adjudications.

Verify in 121–150: Any event-level Held call has a source satisfying the original event criterion; broader causal tests require the stated comparative evidence. No stronger identification requirement is added retrospectively. Type V requires an actually resolved timing error.

Separate last-carried from last-substantively-reviewed dates and explicitly close or hand off every disappearing anomaly ID.

Basis: 097-ie-3, an-091-004-predictive-not-operational, 095-g20-commitment-gap

Tradeoff: A review queue needs attention; not every question receives a daily update.

Verify in 121–150: No ID disappears without closure or successor ID; due reviews cite fresh evidence or a bounded unsuccessful source search. Counts use both distinct threads and appearances.

Use literal read-mode expansions and register a shadow comparison if testing the claimed advantage of O over R/H.

Basis: 092-ie-3, 092-ie-4, 093-ie-2, 093-ie-3, 093-ie-4, 094-ie-3, 094-ie-4, 118-ie-3

Tradeoff: Additional coding and experimental work; the public briefing itself keeps its orientation discipline.

Verify in 121–150: No READ namespace collisions; outcome-blind duplicate coding of a natural sample; no causal mode claim without an external or randomized comparison.

Make research relevance a phenomenon-to-design argument with a rival, a computational intervention and an observable empirical contrast.

Basis: 093-ie-3, 094-ie-2, 095-ie-2, 119-ie-2

Tradeoff: Fewer named connections, but each generates a specific research decision.

Verify in 121–150: Research entries distinguish sourced phenomena from hypotheses, name the study question, explain what an ablation would distinguish and retain disconfirming cases; no claim of a completed study result.

LLM cognitive signatures

No chain carries an explicit prospective S1–S7 assignment. All 113 are unassessable for an observed-failure signature in this audit. This is a missing-data result, not a primary distribution of cognitive failures. Counts for identified S1–S7 are zero; percentages of identified failures are undefined.

SignatureMechanismText or outcome evidence requiredRelated typesIdentified
S1Mean-trajectory regressionTiming drawn toward historical averagesV0
S2Narrative over physical coherenceCompressed inference hides material workV, VII0
S3Plausibility-mass concentrationModal completion excludes other action setsI0
S4Reflexive blindnessActors’ anticipatory adaptation is omittedII0
S5Categorical stickinessA category is retained after its referent changesVI0
S6Buffer blindnessA variable buffer is treated as constantIV0
S7Mechanism–outcome decouplingAny route to the endpoint counts as supportVII0

S1 and S2 cannot be diagnosed from a long horizon alone: the world must first resolve the timing and the material dependency. S3 and S4 require evidence of an omitted action or anticipatory response, not merely an alternative scenario. S5 needs evidence that the category changed, not an unfamiliar noun. S6 requires a demonstrated buffer assumption and its variation. S7 is the strongest construction risk in the mismatched criteria, but the Type VII/S7 resemblance is nearly definitional and must not be counted as independent corroboration. These are coding cautions, not findings about a model’s internals.

Type × Signature, Read-Mode × Type and Read-Mode × Signature contain no observed-failure cells. The YAML stores empty observed cross-tabs and the full unassessable count. Cycle 3’s five S3 assignments cannot be compared with this cycle’s missing prospective labels as if one architecture improved.

Read-mode classification

ModeRetrospective countShare of all chainsHeld / UnresolvedInverted / Spurious / Failed-Z
O11299.1%2 / 1100 / 0 / 0 — not an estimated risk
R00.0%0 / 00 / 0 / 0 — not an estimated risk
H10.9%0 / 10 / 0 / 0 — not an estimated risk

The original token counts are {'O': 104, 'R': 5, 'H': 4}. Seven tokens are semantically invalid for read mode: 092-ie-3, 092-ie-4, 093-ie-2, 093-ie-3, 093-ie-4, 094-ie-3, 094-ie-4. Those raw labels remain in the YAML. Independent coding uses the terminal prose and finds O in those seven cases. In 101-ie-3, three conditional preparation paths earn O despite an original H tag. In 118-ie-3, H is retained as a disclosed boundary case because the author explicitly narrows the terminal test to the final poll/count comparison.

O denotes a retained field of alternatives, not a certificate of good measurement, complete evidence or correct inference. R has no independently coded cases. One H case cannot support a comparison, and classifying most prose O does not show that orienting prevented a severe failure. An external or shadow experimental corpus is still needed.

Complete inference inventory

The complete inventory is searchable through the browser. Each entry includes the source date, original tag, independent stance, test unit, exact chain and its missing outcome evidence. X/Y/Z decomposition, invoked patterns and every source-check mapping are also in the YAML. For recent narrative-format cards, decomposition is interpretive; the exact text remains the authority.

091-ie-1 · 2026-08-17 · O · Unresolved — Three-Body: A Narrow Settlement Can Stabilize One Relation and Destabilize the Whole

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Public shipping and bargaining paths are followed by a CONFIRMS/REFUTES pair about a model comparison. A model result cannot adjudicate what occurred in Hormuz. The title’s triadic analogy is not a mechanism-level connection to entrepreneurial delegation.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: outcomes differ when the same actor preferences are run through clause-level coupling and theater-load constraints. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Equivocality Three-Body: A Narrow Settlement Can Stabilize One Relation and Destabilize the Whole Observation The comprehensive U.S.–Iran settlement deadline has expired while Iran and Oman continue work on a narrower arrangement for Hormuz. Exercises on the Korean Peninsula are being reduced as a U.S. carrier leaves the Pacific for the Middle East. These are separate events, but together they expose an allocation problem: local accommodation can lower immediate friction while redistributing military attention, alliance confidence, and bargaining leverage elsewhere. Inference The useful unit is not “agreement achieved.” It is a bundle of clauses embedded in a theater-load configuration. A shipping clause can improve navigation and reduce near-term strike risk without resolving sanctions, reparations, sovereignty, or the nuclear question. If that local gain also releases pressure for a comprehensive bargain, the same clause has a stabilizing direct effect and a potentially destabilizing indirect effect. Hypothesis In a three-actor field, modular agreements are most durable when their benefits do not sharply reduce one actor’s incentive to maintain the larger bargaining process. Durability should fall when one actor captures the local benefit, another bears the security externality, and the third can gain by delaying the remaining clauses. Release field The narrow settlement can couple back into system repair, contain risk within one channel, or erode leverage for a wider bargain. Release path 1 · Coupled repair Positive tell: a published shipping text is followed by an explicit extension, sequenced nuclear talks, and measurable restoration of traffic. Exclusion tell: shipping resumes while the comprehensive mandate disappears and the military strike cadence rises. Release path 2 · Modular containment Positive tell: Hormuz traffic improves and the local bargain persists even though sanctions and nuclear positions remain unchanged. Exclusion tell: implementation repeatedly depends on progress in the unresolved channels. Release path 3 · Leverage erosion Positive tell: the shipping settlement lowers the cost of delay, talks fragment further, and third-party force allocation changes. Exclusion tell: the narrow accord creates monitoring or enforcement machinery that pulls the larger parties back into regular negotiation. Ripeness Near, across two to fourteen days. The first fair observables are an Iran–Oman text, a formal extension, shipping and insurance data, and changes in strike or carrier posture. CONFIRMS: outcomes differ when the same actor preferences are run through clause-level coupling and theater-load constraints. REFUTES: a binary agreement variable reproduces the same sequence, distribution, and durability of outcomes without clause interactions.
091-ie-2 · 2026-08-17 · O · Unresolved — Epistemic Complementarity: Different Views Matter Only When Their Combination Changes the Discrimination

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: the combined system produces a reproducible interaction gain above the best single agent and equal-resource controls. Bounded component checks are recorded as interaction-tax. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING Ambiguity Epistemic Complementarity: Different Views Matter Only When Their Combination Changes the Discrimination Observation Today’s strongest evidence is split across instruments. COFFIES infers solar activity before it becomes visible but is not ready for real-time operations. NASA’s nebula composite combines observatories that register different wavelengths. Aircraft enter wildfire clouds to sample what satellites see from above. Election observers can assess procedures yet may need local records to reconstruct stolen ballots. No single observer has the whole relevant state. Inference Heterogeneity of perspective is not itself complementarity. It becomes complementarity when combining nonredundant signals changes which hypothesis should be selected or which action should follow, relative to the best single observer and an equal-compute redundant ensemble. Agreement among differently named agents is weak evidence if all consume the same representation or traverse evidence in the same order. Hypothesis Complementarity gains will be largest when agents have differentiated observation functions, partial conditional independence, and an integration rule that preserves disagreement long enough to localize it. Gains will shrink when role labels conceal shared inputs or when coordination cost consumes the information advantage. Release field Differentiated views can produce genuine complementarity, redundant consensus, or coordination loss. Release path 1 · Genuine complementarity Positive tell: the combined system correctly separates cases that the best single agent and equal-compute ensemble confuse, with the gain traceable to a cross-source interaction. Exclusion tell: performance rises only because the multi-agent system receives more tokens, tools, or samples. Release path 2 · Redundant consensus Positive tell: agent outputs converge rapidly, ablation of a role changes little, and apparent confidence rises more than accuracy. Exclusion tell: removing one observation channel selectively destroys performance on the cases that require it. Release path 3 · Coordination loss Positive tell: each agent is locally accurate but integration delays, compresses, or misroutes decisive evidence. Exclusion tell: a lightweight aggregation rule retains the multi-view gain at the same compute budget. Ripeness Immediate for design and near for evidence: one to eight controlled model runs, followed by two to six weeks of ablation and replication. CONFIRMS: the combined system produces a reproducible interaction gain above the best single agent and equal-resource controls. REFUTES: the gain vanishes after resources, input overlap, and search depth are matched.
091-ie-3 · 2026-08-17 · O · Unresolved — AI-Agent GCM: Specialization Is Not Yet an Interaction Mechanism

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: endogenous transitions explain performance after role specialization and compute are controlled. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Complexity AI-Agent GCM: Specialization Is Not Yet an Interaction Mechanism Observation Robin assigns scientific work to specialized agents and produces impressive outputs, but the reported tool-use paths remain strikingly similar in order. The system demonstrates that role division and orchestration can support consequential discovery. It does not by itself establish that adaptive interaction among agents generated the gain. Inference A fixed or nearly fixed pipeline can look multi-agent at the interface while remaining sequential in mechanism. For a GCM of AI agents, specialization, routing, memory, critique, and re-entry should be separate state variables. Otherwise the model risks naming agent plurality while representing only a longer workflow. Hypothesis Endogenous routing will outperform fixed routing only in environments where task uncertainty changes during execution and where agents can detect that change. In stable tasks, fixed sequences may match or beat dynamic coordination because they avoid routing overhead. Release field Agent plurality can yield a specialization gain, an interaction gain, or a coordination burden. Release path 1 · Specialization gain Positive tell: role-specific tools or priors improve local subtask quality, but order perturbations leave final performance stable. Exclusion tell: replacing specialists with matched generalists produces no loss. Release path 2 · Interaction gain Positive tell: agents reroute work after disagreement or new evidence, and those contingent transitions predict success beyond role quality. Exclusion tell: shuffling or freezing the interaction order does not change output quality, novelty, or error recovery. Release path 3 · Coordination burden Positive tell: added agents increase handoffs, duplicated searches, latency, or premature consensus faster than they improve discrimination. Exclusion tell: marginal agents add unique evidence while total resource use is held constant. Ripeness Near, across the next one to twelve benchmark runs once routing, role, and resource controls are independently manipulable. CONFIRMS: endogenous transitions explain performance after role specialization and compute are controlled. REFUTES: a fixed pipeline or single-agent workflow matches the result under equal tools, context, and inference budget.
091-ie-4 · 2026-08-17 · O · Unresolved — Polymathy LLM-ABM: Breadth Counts When It Creates a Bridge, Not When It Enlarges Retrieval

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: cross-domain access creates auditable bridges that alter hypotheses and improve out-of-sample discrimination. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Knightian uncertainty Polymathy LLM-ABM: Breadth Counts When It Creates a Bridge, Not When It Enlarges Retrieval Observation Several signals today invite analogy: hybrid-electric flight preserves unlike energy pathways; multi-instrument astronomy composes unlike wavelengths; protein compression redesigns a structure rather than merely removing pieces. These analogies may generate hypotheses for polymathic agents, but thematic variety alone does not establish useful breadth. Inference Polymathy should be represented as a capacity to discover and test structure-preserving relations across domains. A bridge earns causal status only if it changes the hypothesis space, reveals a discriminating observable, or redirects search. Counting topics, sources, or semantic distance would otherwise reward ornamental breadth. Hypothesis Cross-domain breadth improves discovery when agents retain enough depth to encode constraints from both source and target domains and when a discriminator can reject shallow resemblance. Performance should decline when breadth increases retrieval volume without increasing valid transfers. Release field Cross-domain access can create recombinant bridges, redundant breadth, or overload. Release path 1 · Recombinant bridge Positive tell: a distant-domain relation produces a novel, testable hypothesis that survives checks in both its origin and receiving domains. Exclusion tell: the same hypothesis appears from within-domain retrieval at equal cost. Release path 2 · Redundant breadth Positive tell: topic coverage rises but accepted hypotheses, prediction quality, and search efficiency remain flat. Exclusion tell: removing distant-domain access selectively eliminates high-value hypotheses. Release path 3 · Overload Positive tell: more domains increase false analogies, discriminator burden, and convergence time. Exclusion tell: staged retrieval or stronger source constraints preserve the breadth gain without raising false transfer. Ripeness Near, across five to twenty simulation runs with matched token, retrieval, and evaluation budgets. CONFIRMS: cross-domain access creates auditable bridges that alter hypotheses and improve out-of-sample discrimination. REFUTES: gains reduce to retrieval volume, prompt length, or evaluator preference for diverse language.
092-ie-1 · 2026-08-19 · O · Unresolved — Interfaces Accumulate Load After the Comprehensive System Fails

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: due unadjudicated. The original immediate-to-two-week window ended in early September.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: events cluster on the highest-load interface and removing it changes outcomes across more than one channel. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Three-Body program Interfaces Accumulate Load After the Comprehensive System Fails OBS The Iran settlement clock expired while narrower shipping and mediation routes remained. Two days later, the UAE suspended trade with Iran after missile fire, and Oman itself came under threat. Vessel traffic was already concentrated in an Iran-designated route while Omani lanes carried little recorded traffic. INF The residual interfaces are not independent remnants. They are absorbing functions displaced from the failed comprehensive settlement. That load makes them valuable and visible. Pressure therefore migrates toward the very routes that preserve limited cooperation. HYP In a triadic conflict, a surviving mediator or commercial hub becomes a target when the marginal leverage gained by pressuring it exceeds the coordination value it provides to all parties. Model interface load, not only actor hostility. Release field The loaded interfaces can reconnect the wider system, survive as separate modules, or become the next objects of coercion. Path 1 · Reconnection: Oman and the UAE secure exemptions or texts that reconnect commerce, shipping, and diplomacy. Positive tell: published terms and recovered traffic. Exclusion tell: public meetings without operational volume. Path 2 · Modular survival: one channel remains active while the others stay impaired. Positive tell: stable traffic under narrow rules. Exclusion tell: expanding exemptions that recreate the larger bundle. Path 3 · Bypass capture: coercion moves onto mediators and hubs. Positive tell: retaliation, sanctions, or withdrawal focused on those interfaces. Exclusion tell: temporary rhetoric followed by routine operation. CONFIRMS: events cluster on the highest-load interface and removing it changes outcomes across more than one channel. REFUTES: pressure remains random across targets, or trade and mediation continue unchanged despite threats. Ripeness: immediate to two weeks.
092-ie-2 · 2026-08-19 · O · Unresolved — Complementarity Appears When One Observation Changes the Next Reachable Action

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: information-gain at the handoff predicts collective performance after controlling for agent count and compute. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Epistemic Complementarity Complementarity Appears When One Observation Changes the Next Reachable Action OBS Public astronomers reconstructed a lunar impact trajectory. NASA refined a location. Danuri imaged the crater, and LRO then used the coordinates to produce multi-angle measurements. Self-driving laboratories face the same model-to-instrument-to-update sequence. INF The observers are not complementary merely because they differ. Their relation has work-doing power because each output changes the next observer’s search area, timing, or experiment choice. A handoff is therefore an information transformation tied to action. HYP Group epistemic performance is driven more by the action-value of handoffs than by diversity of agent profiles. The key interaction term is the reduction in the next agent’s feasible search set, net of information lost during translation. Release field The configuration can produce true complementarity, remain a redundant ensemble, or conceal human repair at its interfaces. Path 1 · True complementarity: removing a handoff widens uncertainty or blocks an observation. Positive tell: measurable counterfactual loss. Exclusion tell: unchanged result under ablation. Path 2 · Redundant ensemble: agents provide parallel evidence without changing one another’s action sets. Positive tell: variance reduction but no new reachable test. Exclusion tell: an agent initiates a test unavailable to peers. Path 3 · Repair hidden as complementarity: humans silently correct interface errors. Positive tell: undocumented interventions. Exclusion tell: provenance-complete replay. CONFIRMS: information-gain at the handoff predicts collective performance after controlling for agent count and compute. REFUTES: independent output quality explains all gains. Ripeness: ready for code-level operationalization now.
092-ie-3 · 2026-08-19 · O · Unresolved — A Capability Threshold Can Be Institutional Before It Is Empirically Settled

Originating inference section · Test: comparative mechanism test · Original READ: R

Stance basis: Original READ expansion uses a different meaning from O/R/H. Retrospective terminal-step review finds alternatives retained. Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: safeguard activation changes later estimates even when underlying capability is held constant. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: R · REFLEXIVE AI-Agent GCM A Capability Threshold Can Be Institutional Before It Is Empirically Settled OBS OpenAI says its newest cyber evaluations cannot rule out critical capability and reports responding with stronger safeguards. The source is self-assessment. Independent evidence at the same level is not yet visible in today’s public record. INF Governance can activate on one-sided uncertainty. That may be prudent for tail risk, but it creates a feedback loop: safeguards restrict observation of the very capability they are meant to contain. Later evaluators then see a selected access regime. HYP In agent governance models, threshold rules alter the evidence-generating process. Once controls activate, observed incident rates cannot be interpreted as untreated capability. The model needs potential capability, access, safeguard efficacy, and measurement coverage as distinct variables. Release field Evidence can validate a real threshold, preserve precaution under unresolved uncertainty, or narrow into a measurement enclosure. Path 1 · Real threshold: independent teams reproduce higher end-to-end success on critical tasks. Positive tell: compositional success across environments. Exclusion tell: gains limited to benchmark fragments. Path 2 · Precautionary activation: controls rise while capability remains below the line. Positive tell: unresolved upper confidence bounds. Exclusion tell: direct crossing on replicated tests. Path 3 · Measurement enclosure: access restrictions make external audit impossible. Positive tell: no reproducible protocol or qualified external access. Exclusion tell: independent red-team publication. CONFIRMS: safeguard activation changes later estimates even when underlying capability is held constant. REFUTES: access and evaluation coverage remain stable across the threshold. Ripeness: near, as external evaluations and deployment controls emerge.
092-ie-4 · 2026-08-19 · O · Unresolved — Cross-Domain Translation Is Valuable When It Preserves Constraints

Originating inference section · Test: comparative mechanism test · Original READ: H

Stance basis: Original READ expansion uses a different meaning from O/R/H. Retrospective terminal-step review finds alternatives retained. Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: constraint-preservation predicts transfer success better than semantic similarity alone. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: H · HYPOTHESIS Polymathy program Cross-Domain Translation Is Valuable When It Preserves Constraints OBS A NASA student challenge reframes bridge inspection as an aviation problem. Canada plans to repurpose robotic-arm investment toward a new lunar architecture. DARPA’s current topics place structured reasoning and physically durable memory beside one another. Each case moves an object across a domain boundary. INF Cross-domain movement can produce novelty, but useful recombination must preserve the constraints that made the original system reliable. Aviation cannot ignore civil inspection standards. A lunar arm cannot carry only its hardware and leave acceptance tests behind. Structured reasoning cannot substitute metaphor for an executable representation. HYP Polymathic advantage follows a constrained-transfer function: distance between domains creates novelty, while preserved constraint structure determines whether the transfer works. Very distant combinations with low constraint preservation should generate attractive but brittle proposals. Release field Cross-domain movement can become productive transfer, accumulate translation debt, or remain decorative breadth. Path 1 · Productive transfer: the imported method meets the receiving domain’s acceptance test. Positive tell: certification or benchmark passage. Exclusion tell: only conceptual resemblance. Path 2 · Translation debt: novelty arrives before standards and ownership. Positive tell: repeated manual exceptions. Exclusion tell: stable interface specification. Path 3 · Decorative breadth: multiple domains are named without changing design. Positive tell: ablation leaves the proposal unchanged. Exclusion tell: a borrowed constraint changes the result. CONFIRMS: constraint-preservation predicts transfer success better than semantic similarity alone. REFUTES: domain distance or expertise count explains performance without any interface measure. Ripeness: medium; operational measures can be tested on existing project histories.
093-ie-1 · 2026-08-20 · O · Unresolved — 1 · Tail-first evaluation

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: The hypothesis predicts selection against mean-optimised interventions, but CONFIRMS only asks whether distributional measures qualify a mean. Distributional heterogeneity does not establish the proposed selection mechanism.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: distributional measures reverse or qualify conclusions drawn from means. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 1 · Tail-first evaluation OBS Dense agent networks can converge quickly while producing confident error; aggregate youth outcomes improve while group differences remain; a small share of facilities without waste service reaches 571 million people. INF Mean improvement and tail harm can coexist because system benefits and exposure are unevenly distributed. HYP Interventions optimized for the mean will be selected against once tail outcomes become visible. Release field: healthy convergence; convergence with concentrated failure; apparent convergence created by low-resolution measurement. Path 1 · Mean and tail improve: matched interventions raise central performance and reduce extreme error. Positive tell: quantiles move together. Exclusion tell: the tail is unchanged. Path 2 · Mean improves, tail worsens: coordination benefits typical cases but synchronizes a vulnerable subgroup. Positive tell: higher average performance with more extreme failures. Exclusion tell: tail growth is fully explained by sampling noise. Path 3 · Measurement artifact: the reported mean changes because price, category, or denominator changes. Positive tell: the effect disappears under a stable real-volume measure. Exclusion tell: independent measures reproduce it. CONFIRMS: distributional measures reverse or qualify conclusions drawn from means. REFUTES: matched quantiles, groups, and extreme outcomes all move with the mean. Ripeness: high; existing datasets can report ECDFs and quantiles now.
093-ie-2 · 2026-08-20 · O · Unresolved — 2 · Topology is a treatment, not background

Originating inference section · Test: comparative mechanism test · Original READ: R

Stance basis: Original READ expansion uses a different meaning from O/R/H. Retrospective terminal-step review finds alternatives retained. Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: randomized topology or placement materially changes accuracy, confidence, or recovery after matched resources. Bounded component checks are recorded as interaction-tax. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: R · RESEARCH 2 · Topology is a treatment, not background OBS LLM-agent conformity varies with topology and self-social weighting; violin networks escape local minima by retuning coupling; CONSENSAGENT explicitly intervenes against sycophancy. INF Interaction structure can change the mechanism independently of agent capability. HYP Many apparent “agent quality” effects are topology-by-placement effects. Release field: topology preserves useful heterogeneity; topology accelerates accurate integration; topology amplifies wrong-but-sure cascades. Path 1 · Capability effect: stronger agents outperform across randomized network positions. Positive tell: rank stability survives topology changes. Exclusion tell: performance flips with placement. Path 2 · Topology effect: identical agents yield different distributions under different degree and routing structures. Positive tell: topology changes false-consensus tails under matched message budgets. Exclusion tell: effects vanish once communication volume is held constant. Path 3 · Placement interaction: heterogeneous update rules matter chiefly when assigned to high-degree nodes. Positive tell: random reassignment changes cascade thresholds. Exclusion tell: only population shares matter. CONFIRMS: randomized topology or placement materially changes accuracy, confidence, or recovery after matched resources. REFUTES: agent attributes explain outcomes and topology perturbations add no predictive value. Ripeness: immediate for computational experiments.
093-ie-3 · 2026-08-20 · O · Unresolved — 3 · Re-commitment can reveal task co-evolution

Originating inference section · Test: comparative mechanism test · Original READ: R

Stance basis: Original READ expansion uses a different meaning from O/R/H. Retrospective terminal-step review finds alternatives retained. Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: mechanism freezes change founder-level trajectories and downstream reference points under constant self-access noise. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: R · RESEARCH 3 · Re-commitment can reveal task co-evolution OBS Industrial allocation changes with risk and task conditions; Swift’s rescue separated launch, control, proximity, and capture; institutional actions change later option sets before final outcomes arrive. INF A delegation decision becomes recursive when it alters future task reference points rather than merely assigning the current task. HYP Re-commitment and consultation interact: either can look weak in isolation while their joint sequence shifts tail outcomes. Release field: static routing; adaptive allocation around an unchanged task; recursive allocation that changes the future task. Path 1 · Static routing: performer changes but task state and future options do not. Positive tell: freezing re-commitment leaves trajectories unchanged. Exclusion tell: delegation changes later reference points. Path 2 · Adaptive allocation: allocation responds to state without altering it. Positive tell: current performance improves while future-state distributions remain matched. Exclusion tell: downstream tasks diverge. Path 3 · Co-evolution: consultation plus re-commitment changes what becomes actionable next. Positive tell: a two-by-two freeze produces distributional divergence beyond mean shifts. Exclusion tell: the joint effect is reproduced by added noise or effort. CONFIRMS: mechanism freezes change founder-level trajectories and downstream reference points under constant self-access noise. REFUTES: all differences collapse once effort and information are matched. Ripeness: high for a controlled simulation test.
093-ie-4 · 2026-08-20 · O · Unresolved — 4 · Institutional existence can outlive institutional intelligence

Originating inference section · Test: dated event or document followup · Original READ: H

Stance basis: Original READ expansion uses a different meaning from O/R/H. Retrospective terminal-step review finds alternatives retained. Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: preserved budgets coexist with degraded evaluation, coordination, or tail-event response. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: H · HORIZON 4 · Institutional existence can outlive institutional intelligence OBS CDC programs retain law and funds without staff; a training course can exist without waste infrastructure; a lease process can advance before ecological baselines are settled. INF Institutions can keep producing legible outputs after their sensing and learning capacity is hollowed. HYP The first public symptom will often be distributional drift rather than total program failure. Release field: authority and intelligence remain coupled; outputs persist while learning decays; visible closure follows a long hollowing period. Path 1 · Substitution works: states, grantees, or contractors replace central expertise. Positive tell: evaluation quality and cross-site coordination remain stable. Exclusion tell: variance rises across sites. Path 2 · Hollow persistence: money and reports continue while expertise disappears. Positive tell: outputs arrive but adaptation to new evidence slows. Exclusion tell: staff functions are demonstrably covered. Path 3 · Delayed collapse: operational failure appears only when an unusual event requires tacit knowledge. Positive tell: routine indicators hold until a tail event. Exclusion tell: stress tests show resilient handoffs. CONFIRMS: preserved budgets coexist with degraded evaluation, coordination, or tail-event response. REFUTES: substitute capacity maintains those functions at comparable quality. Ripeness: medium; staffing-function and outcome data are needed.
094-ie-1 · 2026-08-25 · O · Unresolved — 1 · Repricing inside the disclosure window

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: due unadjudicated. The original within-weeks document and behaviour review is due; no exact closing day was specified.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: counterparty behavior tracks terms and ignores announcements. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 1 · Repricing inside the disclosure window OBS A trade deal announced 19 August collapsed 21 August when its terms changed before disclosure; a reduced exercise ended six days early and the counterpart raised its price; a buyback-driven bond rally reversed within one day. INF When an arrangement's terms can move after its announcement, counterparties learn to price the announcement near zero and respond to the terms alone. HYP Repeated repricing shifts counterparties from bargaining to exit, and exit investments persist after the dispute resolves. Release field: restoration with published text; hardened exit and diversification; a standing regime of announcement-layer objects that never acquire terms. Path 1 · Restoration: negotiators return and a text is published. Positive tell: tariff schedules withdrawn against a published agreement. Exclusion tell: talks resume with no text. Path 2 · Hardened exit: the counterpart's diversification becomes contractual. Positive tell: infrastructure starts and trade reallocation within two quarters. Exclusion tell: diversification remains rhetorical while trade shares hold. Path 3 · Announcement regime: both sides keep announcing for domestic audiences while terms never fix. Positive tell: repeated announcement-collapse cycles without schedule changes. Exclusion tell: any operative schedule taking effect. CONFIRMS: counterparty behavior tracks terms and ignores announcements. REFUTES: announcements alone move counterparty concessions durably. Ripeness: high; tariff schedules and Canadian infrastructure decisions are observable within weeks.
094-ie-2 · 2026-08-25 · O · Unresolved — 2 · The content of exchange governs the value of coordination

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: The initial study is evidence that motivated the chain. Re-counting that same study as later confirmation would conflate hypothesis generation with a new test.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: matched-budget content ablations change accuracy and the confidence tail more than topology ablations do. Bounded component checks are recorded as interaction-tax. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 2 · The content of exchange governs the value of coordination OBS Under matched budgets, agent teams sharing complete solutions converge in one round and lose the ensemble's diversity; critique-based exchange helps only when violations are detectable; scale survival of social mechanisms is predictable from structure, and count-versus-percentage encoding moves the threshold. INF Coordination gains and epistemic gains separate: a system can raise agreement while lowering accuracy, and the lever is what is exchanged, not how many exchange it. HYP Systems selected on consensus speed will drift toward full-solution sharing and inherit its confidence tail. Release field: exchange-content design becomes standard; topology remains the default lever and the tax goes unmeasured; consensus metrics get replaced by calibration metrics. Path 1 · Content-aware design: teams share proposals only, or critiques gated by detectability. Positive tell: published ablations crossing content with topology. Exclusion tell: only agent-count and density ablations appear. Path 2 · Tax goes unpriced: production systems keep full sharing for speed. Positive tell: convergence rates rise while accuracy stagnates. Exclusion tell: accuracy tracks convergence. CONFIRMS: matched-budget content ablations change accuracy and the confidence tail more than topology ablations do. REFUTES: content arms are indistinguishable once topology is controlled. Ripeness: high; both papers publish auditable designs.
094-ie-3 · 2026-08-25 · O · Unresolved — 3 · Enforcement recruits the middle layer

Originating inference section · Test: dated event or document followup · Original READ: R

Stance basis: Original READ expansion uses a different meaning from O/R/H. Retrospective terminal-step review finds alternatives retained. Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: compliance moves before penalties land; district policies and third-country banks act ahead of enforcement. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: R · REASONED 3 · Enforcement recruits the middle layer OBS The Iran package defers many secondary sanctions to let third parties "remedy bad behavior"; phone statutes bind districts, not students; the roadless docket's 99%-opposed record is being built for a reviewing court. INF Modern enforcement increasingly lands on intermediaries first, converting them from targets into agents — the deferral, the mandate, and the docket are recruitment devices. HYP Systems governed through recruited intermediaries develop capacity at the middle layer and opacity at the endpoints, because the endpoint behavior stops being directly observed. Release field: effective delegated enforcement; intermediary capture or shirking; endpoint displacement into channels no intermediary sees. CONFIRMS: compliance moves before penalties land; district policies and third-country banks act ahead of enforcement. REFUTES: behavior changes only after direct penalties on endpoints. Ripeness: medium; sanction waivers, district implementation reports, and the final roadless rule will show it within one to two quarters.
094-ie-4 · 2026-08-25 · O · Unresolved — 4 · Downstream products read upstream structure

Originating inference section · Test: comparative mechanism test · Original READ: R

Stance basis: Original READ expansion uses a different meaning from O/R/H. Retrospective terminal-step review finds alternatives retained. Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: helium-4 runs reproduce geometry-specific flow; hiring-threshold changes predict later workforce composition. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: R · REASONED 4 · Downstream products read upstream structure OBS Quark-gluon plasma flow now images the shape of the nucleus that made it; fading black-hole flares encode the captured star's origin; a hiring-filter survey shows AI reallocation happening at admission thresholds that no payroll dataset records. INF Emergent products carry recoverable information about the structures that produced them — and instruments built to observe one direction can be read in the other. HYP Where the forward observable is inaccessible, the backward read becomes the primary measurement; its biases then set what the field believes. Release field: validated backward reads with known bias; backward reads adopted without bias characterization; forward observables recovered and the backward reads retired. CONFIRMS: helium-4 runs reproduce geometry-specific flow; hiring-threshold changes predict later workforce composition. REFUTES: backward inferences fail out-of-sample where forward measurement exists. Ripeness: medium; the helium runs and the next employer-survey wave are scheduled.
095-ie-1 · 2026-08-30 · O · Unresolved — 1 · Milestone separation governs what can be claimed

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: mixed due and future. The card joins a near observation with later comparison or implementation evidence. The near component is reviewable; its entire stated test has a later or unspecified endpoint.

Outcome assessment: The confirming proposition requires a comparative test: downstream measures diverge from early milestone language often enough to change evaluation. Bounded component checks are recorded as nasa-0915. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 1 · Milestone separation governs what can be claimed OBS Roman launched and separated successfully, while commissioning and first science remain ahead; AIxC reported one RoboShare paid order, and the cited records do not supply customer-side confirmation, repeat demand, or unit economics. INF Salient thresholds create pressure to compress a multi-stage process into a binary success claim. HYP Systems that report intermediate milestones separately will calibrate investment and public confidence better than systems that treat first completion as final validation. Release field: downstream validation catches up; early success language persists despite weak downstream performance; or reporting standards begin to separate event, capability, and value thresholds. Path 1 · Validation catches up. Positive tell: Roman completes commissioning and releases calibrated images; RoboShare records repeat customers with disclosed utilization. Exclusion tell: completion claims rise while downstream measures remain absent. Path 2 · Milestone inflation. Positive tell: launch and first-order language is reused as evidence of scientific or market success before those outcomes arrive. Exclusion tell: public reporting consistently labels the unresolved stages. CONFIRMS: downstream measures diverge from early milestone language often enough to change evaluation. REFUTES: early milestones reliably predict later value across comparable programs. Ripeness: near for deployment and repeat-order evidence over weeks; farther for scientific yield over three to six months.
095-ie-2 · 2026-08-30 · O · Unresolved — 2 · Autonomy relocates the scarce human contribution

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: measures of verification labor and exception handling predict adoption outcomes better than autonomy scores alone. Bounded component checks are recorded as arpa-air. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 2 · Autonomy relocates the scarce human contribution OBS ARPA-H targets autonomous stroke intervention through staged validation; DOT offers robotics prizes before field procurement; commercial robot deployments require logistics and safety outside the robot itself. INF As execution automates, verification, authorization, exception recovery, and institutional integration can become more valuable rather than less. HYP Persistent augmentation will be strongest where autonomy expands the verification burden faster than standards can routinize it. Release field: durable human-machine complementarity; transition to routine audited automation; or abdication in which nominal autonomy outruns meaningful verification. Path 1 · Complementarity. Positive tell: autonomy rises while specialist audit hours, escalation judgment, or exception handling remain substantial. Exclusion tell: verification becomes standardized and low-cost. Path 2 · Routine automation. Positive tell: live performance is reliable across settings and human intervention rates fall without a safety penalty. Exclusion tell: edge cases repeatedly return control to scarce experts. Path 3 · Abdication. Positive tell: organizations deploy faster than they can explain, validate, or insure. Exclusion tell: clear authority, halt rules, and public test evidence precede use. CONFIRMS: measures of verification labor and exception handling predict adoption outcomes better than autonomy scores alone. REFUTES: autonomy level predicts substitution without regard to validation architecture. Ripeness: medium; program demonstrations and field pilots unfold over two to five years, while prize and rental evidence arrives sooner.
095-ie-3 · 2026-08-30 · O · Unresolved — 3 · Local knowledge acts when validation rights persist

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: cases with retained validation rights show fewer correction failures and more locally aligned allocation. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 3 · Local knowledge acts when validation rights persist OBS Pacific climate programs connect traditional knowledge to early warning; UNCCD recognizes customary governance; UN Mappers emphasizes co-created data; Nigerian volunteers extend mine-risk education. INF Local knowledge contributes more than coverage when local actors retain the right to validate, correct, or govern downstream use. HYP Programs that transfer data but not validation rights will improve visibility while reproducing dependence. Release field: shared operational authority; consultative inclusion without allocation change; or extraction followed by local disengagement. Path 1 · Shared authority. Positive tell: local priorities alter warning protocols, maps, clearance deployment, finance, or tenure rules. Exclusion tell: final plans cite participation but show no traceable change. Path 2 · Consultation only. Positive tell: recurring events and acknowledgments without correction rights or resource movement. Exclusion tell: contracts, budgets, or governance documents assign local control. CONFIRMS: cases with retained validation rights show fewer correction failures and more locally aligned allocation. REFUTES: centralized systems achieve the same outcomes regardless of local governance rights. Ripeness: near to medium; COP31 texts and mapping outputs can be inspected within months, while adaptation outcomes require longer observation.
095-ie-4 · 2026-08-30 · O · Unresolved — 4 · Observation damage changes the meaning of the number

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: due unadjudicated. The original days-to-weeks revision review is due; the comparative prediction test remains unadjudicated.

Outcome assessment: The confirming proposition requires a comparative test: channel-aware uncertainty predicts revisions or policy errors better than a single confidence range. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 4 · Observation damage changes the meaning of the number OBS Nepal's flood counts moved while communications and access were impaired; jobless claims measured low layoffs but not hiring; a goods-deficit headline combined investment imports with falling exports. INF A number's meaning depends on the observation system and the construct it can actually reach. HYP Decisions improve when uncertainty is attached to measurement channels rather than added as a generic caveat after the estimate. Release field: revised counts converge as access returns; multiple measures separate apparently conflicting mechanisms; or a headline measure continues to dominate despite known construct limits. Path 1 · Convergence. Positive tell: disaster counts stabilize with restored communications and documented reconciliation. Exclusion tell: revisions remain geographically correlated with access gaps. Path 2 · Mechanism separation. Positive tell: payroll, vacancies, trade composition, and final releases distinguish layoffs, hiring, investment, and consumption. Exclusion tell: later measures move together with the headline. CONFIRMS: channel-aware uncertainty predicts revisions or policy errors better than a single confidence range. REFUTES: the headline measures remain stable and decision-sufficient across impaired and normal observation conditions. Ripeness: high for revised disaster and macro releases over days to weeks.
096-ie-1 · 2026-08-31 · O · Unresolved — 1 · A promised retaliation is a disposition, not yet an event

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: due unadjudicated. The original days-to-two-week response window has passed; response form and transit effects remain unadjudicated.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: the response's form (not timing) predicts whether transit volumes and premia move. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 1 · A promised retaliation is a disposition, not yet an event OBS U.S. forces struck launchers on Larak Island on 30 August; Iran's Revolutionary Guards acknowledged casualties and said they would respond; Brent moved to $90.69 by 31 August while 6–8 million barrels daily still transit the strait. INF The configuration is a coerced pause: Iran's mining capability is being attrited before use, which pressures Tehran to respond in a register that restores deterrence without triggering the next strike. HYP The form of the response, more than its occurrence, will set the strait's September regime. Release field: symbolic or proxy response away from the strait; a renewed mining attempt inside it; or a negotiated quiet in which the vow is left rhetorically standing. Path 1 · Response away from the strait. Positive tell: proxy or missile activity against U.S.-linked targets elsewhere while tanker traffic and insurance premia hold. Exclusion tell: naval activity concentrating around Hormuz approaches. Path 2 · Renewed mining. Positive tell: further U.S. strikes on launch or mining assets, insurance premia spiking, transit volumes dropping below the 6 million floor. Exclusion tell: shipping flows and premia stable through the week. Path 3 · Rhetorical settlement. Positive tell: the vow stands unexecuted past mid-September while indirect talks resurface. Exclusion tell: a dated ultimatum or force movements that commit Tehran publicly. CONFIRMS: the response's form (not timing) predicts whether transit volumes and premia move. REFUTES: retaliation occurs with no measurable effect on strait economics, or its absence destabilizes the regime anyway. Ripeness: near — days to two weeks for the response window; the transit-volume regime reads out weekly.
096-ie-2 · 2026-08-31 · O · Unresolved — 2 · The reaction function has been restated; the data have not yet voted

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: due unadjudicated. The original mid-September decision window has passed; the required pre-meeting pricing comparison was not retrieved.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: repricing episodes track data releases more than speeches from here. Bounded component checks are recorded as fed-0916. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 2 · The reaction function has been restated; the data have not yet voted OBS Warsh's 28 August Jackson Hole remarks put inflation first and called financial conditions not broadly restrictive; implied September hike odds rose from ~35% to 60% by 31 August, compounded by the oil move. INF Markets are pricing a chair's stated disposition plus a supply shock, in advance of the inflation and payroll data that would validate either. HYP The odds are a belief barometer that the next two data releases can move by more than any further speech. Release field: data confirm and the hike arrives; data undercut and the odds unwind toward a hold; or the oil shock itself softens demand enough to split the committee visibly. Path 1 · Confirmation. Positive tell: firm inflation prints and resilient payrolls into the September meeting with odds holding above 50%. Exclusion tell: officials publicly walking the speech back. Path 2 · Unwind. Positive tell: soft prints move odds back under 40% within two weeks. Exclusion tell: odds stay elevated through weak data — which would say the market is pricing the person, not the numbers. Path 3 · Visible split. Positive tell: dissent signals or divergent Fed speeches as energy prices pass into headline inflation while growth indicators soften. Exclusion tell: a unanimous September statement either way. CONFIRMS: repricing episodes track data releases more than speeches from here. REFUTES: the September decision lands opposite to the pre-meeting odds without any intervening data surprise. Ripeness: near — bounded by the mid-September meeting window.
096-ie-3 · 2026-08-31 · O · Unresolved — 3 · Machine capability is scaling faster than its verification layer

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: deployment outcomes correlate with verification-layer maturity more than with capability benchmarks. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 3 · Machine capability is scaling faster than its verification layer OBS Unitree listed on 19 August with a first-day surge; D1 was sized to Japanese doorframes on 5 August; trackers report Figure past 1,000 units; the AI Act's provenance-marking regime went into enforcement 2 August with a 2 December boundary for pre-existing systems. INF Both physical and generative machine output are crossing into installed human environments faster than the verification infrastructure — audit labor, insurance, functioning content marks — is being built. HYP Where verification lags, the binding constraint on deployment will surface as integration failure and liability pricing, not capability limits. Release field: verification infrastructure catches up and scaling continues; a visible integration failure reprices the sector; or deployment bifurcates into verified-regulated and unverified-gray channels. Path 1 · Catch-up. Positive tell: insurance products, audit standards, and functioning provenance marks appear at commercial scale by the December boundary. Exclusion tell: compliance deadlines slip or marks prove trivially strippable. Path 2 · Repricing event. Positive tell: a deployed-robot injury, or a mark-evading synthetic-content incident with legal consequence, followed by sector-wide cost increases. Exclusion tell: incidents occur without pricing or regulatory response. Path 3 · Bifurcation. Positive tell: marked and unmarked content, insured and uninsured deployment, coexisting with a measurable price spread. Exclusion tell: one channel absorbs the other within two quarters. CONFIRMS: deployment outcomes correlate with verification-layer maturity more than with capability benchmarks. REFUTES: capability metrics alone predict which deployments succeed. Ripeness: medium — the December compliance boundary is near; the robotics integration evidence accrues over two to six quarters.
096-ie-4 · 2026-08-31 · O · Unresolved — 4 · The numbers moved because the instruments did

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: mixed due and future. The card joins a near observation with later comparison or implementation evidence. The near component is reviewable; its entire stated test has a later or unspecified endpoint.

Outcome assessment: The confirming proposition requires a comparative test: channel-aware uncertainty anticipates revision direction better than a generic confidence band. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 4 · The numbers moved because the instruments did OBS Nepal's missing count jumped from ~2,500 to 4,247 as communications returned; India's GDP beat its own central bank's estimate by nearly a point; Cascadia's slab proved five kilometers shallower than hazard models assumed. INF In each case the revision measures the observation system — restored access, model error, instrument resolution — as much as the world. HYP Attaching uncertainty to the measurement channel, rather than to the estimate, predicts which numbers will move next and in which direction. Release field: counts converge as instruments stabilize; systematic bias is revealed and propagates into dependent models; or headline figures continue to dominate decisions despite known channel damage. Path 1 · Convergence. Positive tell: Nepal's counts stabilize with documented reconciliation; RBI's next round narrows its error. Exclusion tell: revisions stay correlated with access gaps. Path 2 · Propagation. Positive tell: the Cascadia revision enters the National Seismic Hazard Model and repriced insurance within a defined cycle. Exclusion tell: the finding sits outside code and pricing beyond the next revision cycle. CONFIRMS: channel-aware uncertainty anticipates revision direction better than a generic confidence band. REFUTES: headline estimates prove decision-sufficient across damaged and intact observation conditions alike. Ripeness: near for disaster and macro counts; far — years — for hazard-model propagation.
097-ie-1 · 2026-09-01 · O · Unresolved — 1 · The tanker strike's author is the strait's next regime

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: due unadjudicated. The original one-to-two-week attribution window has passed; the joint attribution, premiums and traffic record remains unadjudicated.

Outcome assessment: The confirming proposition requires a comparative test: war-risk premia and transit volumes move with the attribution state rather than the incident count. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 1 · The tanker strike's author is the strait's next regime OBS Iran struck U.S. bases in Jordan and sent a drone at the UAE, 30–31 August, claiming retaliation; a tanker in the Hormuz southern lane was hit by three unclaimed projectiles early 1 September per UKMTO; Pezeshkian offered reciprocity on the Islamabad memorandum from Bishkek the same day. INF The response decomposed into an owned channel calibrated for deterrence credit and an unowned incident whose authorship determines whether the strait itself is back in play. HYP Attribution — not incident count — sets the September shipping regime. Release field: attribution lands on Iran or a proxy; no attribution and no repetition; or U.S. escalation on the standing "hit them hard" vow regardless of authorship. Path 1 · Attributed to Iran or a proxy. Positive tell: CENTCOM or UKMTO attribution, war-risk premia spiking, follow-on strikes on maritime assets. Exclusion tell: official silence past a week with premia mean-reverting. Path 2 · Unattributed and unrepeated. Positive tell: lanes quiet through mid-September, premia settling, the Bishkek offer developing into talks. Exclusion tell: a second incident in the lanes, attributed or not. Path 3 · U.S. escalation regardless. Positive tell: strike packages against Iranian maritime or launch assets within days, force flow into the theater. Exclusion tell: the vow standing rhetorical through mid-September while the offramp is explored. CONFIRMS: war-risk premia and transit volumes move with the attribution state rather than the incident count. REFUTES: strait economics reprice materially with no attribution and no further incidents — ambiguity alone doing the work. Ripeness: near — attribution windows in this waterway have historically resolved or gone permanently cold within one to two weeks.
097-ie-2 · 2026-09-01 · O · Unresolved — 2 · The ECB inherits the Fed's dilemma without the Fed's pre-commitment

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: due unadjudicated. The September ECB decision and pricing window has passed; comparative curve evidence is missing.

Outcome assessment: The confirming proposition requires a comparative test: the euro curve's September path is predicted better by the energy component than by core. Bounded component checks are recorded as ecb-0910. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 2 · The ECB inherits the Fed's dilemma without the Fed's pre-commitment OBS Euro-area flash inflation printed 3.3% for August on 1 September, energy at +14.3%, while core fell to 2.4%; country dispersion runs from Estonia's 1.3% to Lithuania's 5.8%; CNBC's same-day analysis expects higher rates to follow. INF The euro area is absorbing the same imported energy shock the U.S. curve repriced last week, but with no stated reaction function anchoring expectations and a wider internal dispersion to govern. HYP Euro money-market pricing over two weeks will reveal whether markets price the shock or the institution. Release field: hike expectations firm on the energy pass-through; the ECB signals look-through and pricing follows core down; or dispersion stress dominates and the policy debate turns to fragmentation rather than level. Path 1 · Pass-through pricing. Positive tell: euro money markets adding tightening within two weeks, ECB speakers foregrounding headline. Exclusion tell: officials uniformly framing the print as imported and temporary. Path 2 · Look-through. Positive tell: pricing tracking the core decline, emphasis on the 2.4% figure in official communication. Exclusion tell: a second energy-driven headline jump in September with core no longer falling. Path 3 · Fragmentation foreground. Positive tell: peripheral-core spread language returning to ECB commentary as Spain's 4.5% diverges further from France's 2.7%. Exclusion tell: dispersion narrowing as the energy shock distributes evenly. CONFIRMS: the euro curve's September path is predicted better by the energy component than by core. REFUTES: pricing follows core downward while headline and energy keep climbing. Ripeness: near — bounded by the ECB's next scheduled decision and two intervening data prints.
097-ie-3 · 2026-09-01 · O · Unresolved — 3 · The forum question: does the universal table still assign action?

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: The retrieved chair’s statement contains dissent and action language. The archive’s no-text premise requires correction; the longer claim about forum authority still needs implementation and venue-following evidence.

Timing: mixed due and future. The card joins a near observation with later comparison or implementation evidence. The near component is reviewable; its entire stated test has a later or unspecified endpoint.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: forum outputs sort onto the assignment/preservation axis and predict where the next live subject surfaces. Bounded component checks are recorded as g20-0901. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 3 · The forum question: does the universal table still assign action? OBS The G20 ministerial's window closed 1 September with no communiqué located by this cutoff, against Bessent's growth-diagnosis framing; the SCO summit opened the same day in Bishkek and immediately hosted the Hormuz conversation's most consequential public statement. INF The multilateral field is bifurcating into a universal table that diagnoses and bloc tables that convene the live subjects — form persisting at one, work migrating to the other. HYP The presence or absence of assignment language in whatever G20 text emerges is a dated test of the universal forum's residual work-doing power. Release field: a communiqué with named commitments lands; a chair's statement or nothing preserves the forum without assignment; or the December leaders' summit inherits an agenda the ministerial declined to bind. Path 1 · Assignment. Positive tell: text with commitments, timetables, or reporting mechanisms within days. Exclusion tell: the window closing on remarks and photographs alone. Path 2 · Preservation. Positive tell: a chair's summary cataloguing "discussions" without dispositive verbs. Exclusion tell: any dissent note — dissent implies there was a text worth disputing. Path 3 · Deferral to Miami. Positive tell: officials framing December's leaders' summit as the decision venue. Exclusion tell: the Miami agenda forming around new subjects rather than the ministerial's leftovers. CONFIRMS: forum outputs sort onto the assignment/preservation axis and predict where the next live subject surfaces. REFUTES: a substantive assigned-action text emerges and the subject migration to bloc venues reverses. Ripeness: near for the text verdict — days; far for the bifurcation claim, which reads out across the December summits.
097-ie-4 · 2026-09-01 · O · Unresolved — 4 · The instruments are converging, each on its own clock

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: mixed due and future. The card joins a near observation with later comparison or implementation evidence. The near component is reviewable; its entire stated test has a later or unspecified endpoint.

Outcome assessment: The confirming proposition requires a comparative test: instruments with dated review points show smaller terminal revisions than instruments without them. Bounded component checks are recorded as fed-0916. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 4 · The instruments are converging, each on its own clock OBS Nepal's missing count fell from 4,247 to 3,916 as districts reconciled while the dead rose past 1,000 — the documented-reconciliation tell yesterday's chain named for its convergence path; the U.S. measles elimination review is scheduled for November against 2,903 cases; the eighteen-point futures–Kalshi spread awaits its 16 September resolution. INF Each observation system carries a dated moment when its reading gets scored against the world — some self-scheduled, some imposed. HYP Systems with scheduled scoring converge; systems without it drift until an event forces the reconciliation. Release field: scheduled tests arrive and the instruments converge; a test is deferred and the drift compounds; or an unscheduled event scores an instrument that had no review date. Path 1 · Scheduled convergence. Positive tell: Nepal's missing count continuing to fall with documented reconciliation; the November measles review and the September FOMC arriving on time and being scored publicly. Exclusion tell: review dates slipping quietly. Path 2 · Unscheduled scoring. Positive tell: an event — a late hurricane, an attributed strike — repricing an instrument that carried no review date. Exclusion tell: the quiet season ending quietly everywhere at once. CONFIRMS: instruments with dated review points show smaller terminal revisions than instruments without them. REFUTES: revision size proves independent of whether a scoring date existed. Ripeness: near for Nepal and the FOMC spread — days to two weeks; medium for the November measles review; far for hazard-model propagation, carried from yesterday.
098-ie-1 · 2026-09-03 · O · Unresolved — State-linked ownership either acquires control or retreats to permission

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: public agreements disclose enforceable rights in Venezuela and closed financing with commissioned route segments in Nigeria. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING State-linked ownership either acquires control or retreats to permission OBS Washington announced an equity stake in private oil firms operating in Venezuela; Chevron and Eni signed expansion agreements in Caracas on 2 September. Nigeria separately reports $845 million in commitments toward a $2 billion fibre plan and a $750 million cloud-investment target. INF Both cases use state-linked capital language to convert strategic intent into an investable structure. The configurations differ: Venezuela concerns ownership and sanctions around existing resources, while Nigeria concerns finance and construction for new infrastructure. HYP The common disposition is toward thicker state participation where commercial risk is difficult to price, but the form may release as control, catalytic support, or promotional language. Release field Control-rights path. Positive tell: disclosed voting, approval, or exclusion rights attach to the Venezuelan stake. Exclusion tell: the agreements preserve company discretion and grant no continuing state authority. Catalytic-finance path. Positive tell: Project BRIDGE reaches financial close and private co-finance follows the public commitments. Exclusion tell: commitments remain unsigned or construction does not begin. Scope-retreat path. Positive tell: later documents narrow equity or kilometre claims while preserving the original headline. Exclusion tell: implementation matches the announced scope. Ripeness: near for documentary clarification, within weeks to months; farther for physical delivery, over quarters. CONFIRMS: public agreements disclose enforceable rights in Venezuela and closed financing with commissioned route segments in Nigeria. REFUTES: the stake proves purely nominal and the infrastructure targets advance without any material state-linked control or finance.
098-ie-2 · 2026-09-03 · O · Unresolved — Control systems will be judged at the failure interface

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: disclosed testing and audit data show improvement at the treated interface alongside identifiable residual failure modes. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Control systems will be judged at the failure interface OBS OpenAI described automated shutdown work after a safety-test escape. Indonesia suspended a Sidoarjo kitchen after a food-safety incident, while more than 800 people were reported ill across three incidents during the week. INF A shutdown control and a meal-safety regime are unlike technologies, yet both are protective systems whose value appears only when a process departs from normal operation. Post-incident action establishes response capacity, not prevention. HYP Each system will initially improve the visible control while leaving a less visible dependency—trigger quality, permissions, vendor monitoring, or laboratory feedback—as the next constraint. Release field Preventive-learning path. Positive tell: later tests or audits identify failures before escape or illness. Exclusion tell: intervention remains entirely post-incident. Control-displacement path. Positive tell: the new safeguard works, but failures move to uncovered systems, vendors, or triggers. Exclusion tell: coverage expands without displacement in repeated tests. Narrow-fix path. Positive tell: one kitchen or one failure mode is addressed while comparable incidents recur elsewhere. Exclusion tell: system-wide rates fall against a valid comparison. Ripeness: near for test scope and kitchen audits, within one to three months; farther for reliable incidence trends. CONFIRMS: disclosed testing and audit data show improvement at the treated interface alongside identifiable residual failure modes. REFUTES: independent evidence shows the controls already cover the relevant system comprehensively and prevent recurrence.
098-ie-3 · 2026-09-03 · O · Unresolved — Overshoot governance separates metric return from system recovery

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: policy instruments publish distributional metrics, explicit removal assumptions, and named responsibility for irreversible harms. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Overshoot governance separates metric return from system recovery OBS UNEP's 2 September report examines pathways in which warming exceeds 1.5°C, peaks, and later declines. A 3 September report describes weaker Indian monsoons under El Niño alongside heavier rainfall extremes. INF Both observations show why scalar targets conceal distributions. A global temperature can later fall while particular losses persist; a seasonal monsoon can weaken while local downpours intensify. HYP Institutions will increasingly govern tail distributions and irreversible components rather than relying on a single mean target, but only where measurement and authority are assigned at the relevant spatial and temporal scales. Release field Adaptive-metrics path. Positive tell: policies report duration above 1.5°C, irreversible losses, and local rainfall tails alongside the mean. Exclusion tell: official reporting remains centred on one headline average. Removal-dependence path. Positive tell: overshoot pathways become contingent on specified carbon-removal scale and timing. Exclusion tell: credible pathways avoid material reliance on later removal. Fragmented-response path. Positive tell: local adaptation advances without an institution able to coordinate the aggregate transition. Exclusion tell: mandates and finance are assigned across levels. Ripeness: near for pathway assumptions and reporting choices; far for evidence of climatic return, across years to decades. CONFIRMS: policy instruments publish distributional metrics, explicit removal assumptions, and named responsibility for irreversible harms. REFUTES: a single temperature or seasonal-rainfall measure proves sufficient to predict the relevant damages and responses.
098-ie-4 · 2026-09-03 · O · Unresolved — Formal inclusion will be tested by allocation and participation

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: the first consultations alter allocation or implementation, and matched school data reveal whether exclusion burdens change. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING Formal inclusion will be tested by allocation and participation OBS New Zealand's disability-support legislation creates a legal framework and consultation requirements without changing current entitlements. Two reports drawing on Arbor school data describe different three-year trends in England’s internal exclusions; the newer report also finds disproportionate burdens on disadvantaged pupils and pupils with special educational needs and disabilities. INF Both developments separate formal presence from substantive participation. A consultation right can exist without changing resources; a pupil can remain enrolled while excluded from ordinary classroom activity. The school trend itself also changes with the observation frame. HYP The legitimacy of inclusion frameworks will turn on whether voice changes allocations and whether nominal membership preserves access to the core service. Release field Substantive-inclusion path. Positive tell: consultation changes rules, funding, or service quality and matched-panel school exclusion duration falls. Exclusion tell: process expands while access remains unchanged. Formalization path. Positive tell: law and reporting improve consistency without altering entitlements or classroom participation. Exclusion tell: measurable distributional effects emerge. Selective-burden path. Positive tell: costs remain concentrated among disabled, poorer, or otherwise less powerful participants. Exclusion tell: disparities narrow after controlling for coverage and denominator changes. Ripeness: near for commencement and definitions; medium, over one school and budget cycle, for distributional effects. CONFIRMS: the first consultations alter allocation or implementation, and matched school data reveal whether exclusion burdens change. REFUTES: process measures show no relationship to participation, access, or distribution across groups.
099-ie-1 · 2026-09-08 · O · Unresolved — 1 · Portfolio value depends on coupling rights

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: public documents show cross-component authority and at least one decision altered through it. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 1 · Portfolio value depends on coupling rights OBS The EU–Greenland package joins eight domains under a declaration and €200 million envelope; UNESCO joins AI, literacy, sovereignty, and education as a common good across one meeting week. INF Bundling creates value only when information, authority, and resources cross component boundaries without one component silently governing the rest. HYP Over the next three to twelve months, the portfolios lean toward either operational coupling or channel decomposition. Path A · Operational coupling. Positive tell: named projects share budgets, local approval, measures, and accountable owners. Exclusion tell: sectors receive separate announcements with no common decision rule. Path B · Dominant-channel capture. Positive tell: extraction, procurement, or vendor access determines priorities across the package. Exclusion tell: service obligations and community or teacher rights can block or redirect implementation. CONFIRMS: public documents show cross-component authority and at least one decision altered through it. REFUTES: implementation proceeds as unrelated projects whose common framing has no observable effect.
099-ie-2 · 2026-09-08 · O · Unresolved — 2 · Space capital changes cadence only through test throughput

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: capital arrival predicts a measurable increase in completed tests and repeat mission performance. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 2 · Space capital changes cadence only through test throughput OBS Isar delivered five satellites and an experiment to orbit on 5 September; Stoke announced a $1 billion initial Series E close on 8 September. INF Finance can buy parallel tests, manufacturing, and failure tolerance, while orbital delivery verifies an integrated chain that financing cannot. HYP The release is near enough to observe over six to eighteen months, but the evidence can separate into two paths. Path A · Learning acceleration. Positive tell: funded amount is followed by more test articles, transparent failures, and shorter intervals between milestones. Exclusion tell: spending rises while milestone dates recede or become less specific. Path B · Narrative thickening. Positive tell: valuation and capacity claims expand faster than independent hardware evidence. Exclusion tell: repeat flights and customer payload delivery establish a stable cadence. CONFIRMS: capital arrival predicts a measurable increase in completed tests and repeat mission performance. REFUTES: test throughput and milestone resolution do not improve after financing.
099-ie-3 · 2026-09-08 · O · Unresolved — 3 · Aggregate weakness releases through labor or investment first

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: mixed due and future. The card joins a near observation with later comparison or implementation evidence. The near component is reviewable; its entire stated test has a later or unspecified endpoint.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: the next official releases preserve the same ordering among employment, investment, production, and demand. Bounded component checks are recorded as nbs-1008. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · ORIENTING 3 · Aggregate weakness releases through labor or investment first OBS China's August PMI remained 49.8 while production and orders exceeded 50 and employment stood at 48.7; South African GDP fell 0.2% in the second quarter amid weak investment and several contracting sectors. INF Headline output can conceal which adjustment channel is carrying pressure. Firms may preserve production by thinning labor or investment before output falls further. HYP Across the next one to two data releases, the configuration can stabilize or cascade. Path A · Thin stabilization. Positive tell: output and orders hold above threshold while employment or investment stays weak. Exclusion tell: hiring, fixed investment, and household demand recover together. Path B · Broader cascade. Positive tell: weakness spreads from labor and investment into production and demand. Exclusion tell: revisions or third-quarter data reverse the contraction without sectoral propagation. CONFIRMS: the next official releases preserve the same ordering among employment, investment, production, and demand. REFUTES: the subindices converge through a broad recovery or a one-off revision.
099-ie-4 · 2026-09-08 · O · Unresolved — 4 · Ecological intervention succeeds when it preserves the timing relation

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: deadline not operationalized. A proposed run count, readiness statement or qualitative horizon does not supply a scheduled final observation. It cannot be classified as an expired forecast solely from elapsed calendar time.

Outcome assessment: The confirming proposition requires a comparative test: replication identifies the same relational variable and shows that changing it changes outcomes. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · ORIENTING 4 · Ecological intervention succeeds when it preserves the timing relation OBS Coral priming depends on prior exposure; sponge metamorphosis depends on sunset-associated cues; Japan's amended ordinance operates against invasive-species spread; Mekong farm outcomes are reported across cost and profit. INF Each case makes sequence and environment part of the mechanism. Acting on the organism, statute, or farm practice alone may miss the relation that produces the outcome. HYP Evidence will mature over one to three field seasons through replication and implementation. Path A · Relation-preserving intervention. Positive tell: effects replicate across environments when exposure, cues, enforcement timing, or farm sequence are specified. Exclusion tell: results vanish when moved outside the original controlled configuration. Path B · Context-bound effect. Positive tell: outcomes depend on one species, laboratory condition, jurisdiction, or selected participant group. Exclusion tell: independent field sites reproduce the mechanism and distribution of effects. CONFIRMS: replication identifies the same relational variable and shows that changing it changes outcomes. REFUTES: the reported effects are explained by uncontrolled correlates or selection.
100-ie-1 · 2026-09-11 · O · Unresolved — Corridor control becomes leverage only when dependent actors respond

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Retention paired with observable commercial or naval adaptation. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: NEAR, 2–6 WEEKS Corridor control becomes leverage only when dependent actors respond OBSERVATION Houthi forces reportedly captured Mokha on 10 September and Mayun on 11 September; Saudi Arabia struck the port city’s airport, and the Security Council convened after fresh attacks. INTERPRETATION The configuration creates a disposition toward higher shipping and security costs, but contested territory is not equivalent to durable control. HYPOTHESIS Retention could let limited forces impose disproportionate costs by changing expectations at the Bab el-Mandeb approach. Release field Commercial activation. Positive tell: vessels reroute or insurers change terms while control persists. Exclusion tell: traffic and terms remain stable despite retention. Military containment. Positive tell: counterattack removes operating capacity before route behavior changes. Exclusion tell: repeated strikes fail to dislodge control. Symbolic seizure. Positive tell: territorial claims continue but no sustained operations appear. Exclusion tell: logistics, fortification, or repeat interdiction becomes visible. CONFIRMS: Retention paired with observable commercial or naval adaptation. REFUTES: Rapid loss of the sites with no measurable route, insurance, or deployment response.
100-ie-2 · 2026-09-11 · O · Unresolved — New measurement capacity can move the bottleneck from sensing to verification

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: Verification queues, external tests, or decision thresholds become the limiting resource. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: MEDIUM, 1–3 QUARTERS New measurement capacity can move the bottleneck from sensing to verification OBSERVATION EPA is screening for more than 1,000 contaminants; a virtual-cell model uses more than 38 million measurements; women’s-health researchers are proposing broader real-world and citizen evidence. INTERPRETATION Each system enlarges the set of detectable associations faster than it necessarily enlarges the capacity to judge causal or practical importance. HYPOTHESIS Institutions will gain the most when they reserve verification resources and define downstream decision rules before signal volume expands. Release field Verified expansion. Positive tell: external validation changes regulations, trials, or laboratory choices. Exclusion tell: signal volume rises without changed decisions. Pseudo-resolution. Positive tell: dashboards and models proliferate while uncertainty about action remains. Exclusion tell: explicit thresholds reliably trigger review and intervention. Selective retreat. Positive tell: programs narrow to a smaller set of reproducible signals. Exclusion tell: collection expands without prioritization. CONFIRMS: Verification queues, external tests, or decision thresholds become the limiting resource. REFUTES: Expanded measurement produces reliable downstream decisions without material checking constraints.
100-ie-3 · 2026-09-11 · O · Unresolved — Redress changes systems only when claims feed operating decisions

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Final rules connect claims data to monitoring, enforcement, and corrective plans. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: MEDIUM, 3–12 MONTHS Redress changes systems only when claims feed operating decisions OBSERVATION ORR consulted on fair redress for failures of booked assistance while wider rail reform moved through licence, monitoring, and enforcement design. INTERPRETATION Compensation can recognize harm, but only aggregated failure data and enforceable responsibility can alter the service that produced it. HYPOTHESIS Automatic redress plus transparent failure reporting would couple passenger experience to operator incentives more strongly than case-by-case claims. Release field Preventive feedback. Positive tell: claims are aggregated, disclosed, and followed by falling comparable failures. Exclusion tell: compensation rises with no operational response. Priced exclusion. Positive tell: payments standardize while assistance reliability remains flat. Exclusion tell: bookings and successful assistance rise together. Administrative friction. Positive tell: evidence burdens suppress valid claims. Exclusion tell: automatic triggers remove most passenger effort. CONFIRMS: Final rules connect claims data to monitoring, enforcement, and corrective plans. REFUTES: The framework governs payment calculations only and produces no service-level feedback.
100-ie-4 · 2026-09-11 · O · Unresolved — Shared technical infrastructure creates capacity when access produces cumulative practice

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Repeated cross-institution use produces validated workflows and durable collaborations. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: FAR, 1–3 YEARS Shared technical infrastructure creates capacity when access produces cumulative practice OBSERVATION Jülich placed JION inside a hybrid research infrastructure, while ESSERC convened European device, circuit, and system-on-chip communities. INTERPRETATION Access and convening can build problem-selection, integration, and verification competence before frontier performance becomes decisive. HYPOTHESIS Europe’s near-term technical advantage may depend more on cumulative hybrid practice than on stand-alone hardware announcements. Release field Cumulative learning. Positive tell: repeat users, shared workflows, publications, and funded collaborations grow. Exclusion tell: access remains demonstration-heavy. Infrastructure theater. Positive tell: announcements outpace usage and reproducible results. Exclusion tell: external users report workload gains. Gatekeeper concentration. Positive tell: access expands formally but a narrow group controls allocation and verification. Exclusion tell: broad, transparent allocations produce diverse use. CONFIRMS: Repeated cross-institution use produces validated workflows and durable collaborations. REFUTES: The facilities remain lightly used or add no contribution beyond classical resources and ordinary conferences.
101-ie-1 · 2026-09-14 · O · Unresolved — Broader coordination produces action only when a member accepts a costly transfer

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Named resources, authority, or behavior move through the new arrangements. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: NEAR, 2–8 WEEKS Broader coordination produces action only when a member accepts a costly transfer OBSERVATION BRICS adopted a broad declaration; ASEAN and China prepared a working visit; the African Union convened an arms-amnesty event; EU officials maintained sanctions instruments. INTERPRETATION Each arrangement creates a channel, but the channels carry different combinations of language, access, resources, legal authority, and personal risk. HYPOTHESIS The most consequential coordination will first appear where a named actor accepts a duty or cost that cannot be explained by ceremony alone. Release field Operational coupling. Positive tell: a budget, deadline, legal instrument, collection result, or working body changes conduct. Exclusion tell: communiqués repeat without assigned implementation. Modular alignment. Positive tell: cooperation advances in selected domains while disagreement remains explicit elsewhere. Exclusion tell: one dispute blocks every channel. Declarative maintenance. Positive tell: relationships remain visible but no costly transfer occurs. Exclusion tell: implementers acquire specific duties. CONFIRMS: Named resources, authority, or behavior move through the new arrangements. REFUTES: The next cycle reproduces the language without an accountable handoff.
101-ie-2 · 2026-09-14 · O · Unresolved — Interface-rich technical systems outperform component-rich systems when mismatch is observable

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: Independent evaluations show lower downstream error or faster recovery attributable to interface controls. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: MEDIUM, 1–4 QUARTERS Interface-rich technical systems outperform component-rich systems when mismatch is observable OBSERVATION High-NA mask development, a lunar foundation model, an enterprise AI harness, linked tissue models, T-wave warning research, and staged surgical autonomy all foreground transfers between components. INTERPRETATION The common configuration is coupled capability. Performance depends on whether the receiver can interpret the sender’s state, constraints, and uncertainty. HYPOTHESIS Systems with explicit provenance, authority boundaries, and recovery tests will deliver more reliable downstream work than systems optimized primarily for isolated benchmark performance. Release field Inspectable coupling. Positive tell: interface failures are detected early and corrected without system-wide loss. Exclusion tell: teams cannot locate which transfer failed. Administrative drag. Positive tell: added controls slow work without lowering consequential errors. Exclusion tell: error cost and recovery time decline. Hidden component dominance. Positive tell: one model or device remains the main source of failure despite strong interface design. Exclusion tell: residual failures concentrate at handoffs. CONFIRMS: Independent evaluations show lower downstream error or faster recovery attributable to interface controls. REFUTES: Component capability alone explains performance, or interface layers add cost without measurable protection.
101-ie-3 · 2026-09-14 · O · Unresolved — High-confidence forecasts create value when they trigger reversible preparation

Originating inference section · Test: comparative mechanism test · Original READ: H

Stance basis: The original H tag is not earned by the terminal prose: conditional preparation, theatre and regional divergence remain live. Independently coded O; no outcome information enters this decision.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: Precommitted triggers lead to timely, revisable action as local signals arrive. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: H · RIPENESS: NEAR, 1–6 MONTHS High-confidence forecasts create value when they trigger reversible preparation OBSERVATION NOAA assigns very high probability to a very strong El Niño, while recent climate and pollution findings show that regional effects follow different pathways and can arrive as pulses. INTERPRETATION A global forecast does not specify a local impact. Its practical value lies in opening a window for conditional preparation before regional evidence becomes decisive. HYPOTHESIS Organizations that predefine local triggers and low-regret actions will capture forecast value without treating uncertainty as certainty. Release field Conditional preparation. Positive tell: agencies stage resources and publish thresholds for deployment. Exclusion tell: plans wait for damage or commit irreversibly too early. Forecast theater. Positive tell: confident public language appears without operational changes. Exclusion tell: budgets, staffing, or protocols move. Regional divergence. Positive tell: local indicators redirect preparation away from the global modal story. Exclusion tell: impacts track the dominant pathway broadly. CONFIRMS: Precommitted triggers lead to timely, revisable action as local signals arrive. REFUTES: Preparations cannot be linked to the forecast or create larger costs than waiting.
101-ie-4 · 2026-09-14 · O · Unresolved — Open infrastructure compounds only when communities can carry context across provider boundaries

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Teams preserve usable context across the OSF transition and Artemis data are reused across institutional boundaries. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: FAR, 1–3 YEARS Open infrastructure compounds only when communities can carry context across provider boundaries OBSERVATION Artemis partners are building data-sharing practices while OSF Projects is moving toward read-only status under funding pressure. INTERPRETATION Access to files is only one layer of openness. Cumulative work also requires provenance, relationships, permissions, stable references, and funded stewardship. HYPOTHESIS The durability of an open-science commons will depend on whether teams can migrate context, not just download content. Release field Portable commons. Positive tell: migrations retain identifiers, histories, permissions, and reproducibility. Exclusion tell: only raw files survive. Archive without practice. Positive tell: public records remain accessible while active coordination fragments. Exclusion tell: successor services preserve workflows. Stewardship renewal. Positive tell: funders treat maintenance and interoperability as explicit research infrastructure. Exclusion tell: support remains project-by-project and temporary. CONFIRMS: Teams preserve usable context across the OSF transition and Artemis data are reused across institutional boundaries. REFUTES: Static access alone proves sufficient for cumulative work.
102-ie-1 · 2026-09-15 · O · Held · stated event criterion — Depleted buffers force a choice between restriction and deadline slip

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: adjudicated event. The stated event/document criterion is already met within its prospective window.

Outcome assessment: Held at the original event criterion: the Environment Agency records a rise in abstraction restrictions before 31 October while storage continued to decline. The additional dated enforcement position is a separate institutional response. The broader cross-case causal explanation is untested, and no general mechanism validation is claimed.

READ: O · RIPENESS: NEAR, 3–10 WEEKS Depleted buffers force a choice between restriction and deadline slip OBSERVATION England’s drought status, the Central European gas contract gap, US topsoil and fire ratios, and UK vacancies all report margins that are gone or going. INTERPRETATION Each margin is governed by a fixed calendar (winter recharge, the January 2027 LNG ban, the October planting window, the end of the fire season) that does not move with the margin. HYPOTHESIS Institutions will reveal their priorities by which they move first: the restriction or the date. Release field Restriction escalation. Positive tell: drought orders, demand-reduction mandates, or planting adjustments are issued. Exclusion tell: deadlines are extended instead. Deadline slip. Positive tell: the pipeline ban moves to 1 November or later, or drought-group timelines defer. Exclusion tell: dates hold while restrictions bite. Recharge arrives. Positive tell: autumn rain restores storage, or long-term supply contracts are signed. Exclusion tell: storage flat into November and no contracts reported. CONFIRMS: At least one named institution changes either its restriction level or its date by the end of October. REFUTES: Margins recover without institutional change, or institutions hold both restriction and date while the physical quantity keeps falling.
102-ie-2 · 2026-09-15 · O · Unresolved — An energy shock transmits through producer prices before it reaches demand

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Consumer-price and vacancy series move in the direction one path predicts across at least two economies. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: NEAR, 1–3 MONTHS An energy shock transmits through producer prices before it reaches demand OBSERVATION India’s fuel-and-power inflation at 22.93%, New York prices paid at 63.1, a euro-area surplus €76 billion smaller year to date, and Gulf LNG exports down 43 billion cubic metres. INTERPRETATION The shock has reached input costs and trade balances; demand indicators (UK and euro-area vacancies, China’s retail sales) are weakening for reasons the releases do not attribute. HYPOTHESIS The next quarter will show whether input-cost pressure passes through to consumer prices or is absorbed by margins under weak demand. Release field Pass-through. Positive tell: consumer inflation accelerates in India, the United States, and the euro area within two releases. Exclusion tell: core measures stay flat. Margin absorption. Positive tell: producer prices rise while consumer prices and employment fall, and profit warnings appear. Exclusion tell: corporate margins hold. Demand destruction. Positive tell: volumes fall faster than prices rise and vacancies keep falling. Exclusion tell: hiring stabilizes. CONFIRMS: Consumer-price and vacancy series move in the direction one path predicts across at least two economies. REFUTES: The producer-price signal reverses within two months, marking a short pulse rather than a transmission.
102-ie-3 · 2026-09-15 · O · Unresolved — Commissioned instruments will be judged by whether results arrive on the stated clock

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: At least three of the six instruments named above report a dated result or certification within their stated windows. Bounded component checks are recorded as nasa-0915. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · RIPENESS: FAR, 1–3 YEARS Commissioned instruments will be judged by whether results arrive on the stated clock OBSERVATION A muonium beam, Roman’s camera, FLEX and Sentinel-3C, IBM’s error correction toolkit, Digit 5, and the Xe-100 assessment each carry a stated result window. INTERPRETATION The structural risk is not in the instruments but in the reading, when a commissioned instrument is counted as a result. HYPOTHESIS Instruments whose owners publish result calendars and independent-validation plans will produce usable evidence; those that do not will produce announcements. Release field Results on schedule. Positive tell: first data, certification, or a permit arrives within the stated window with third-party involvement. Exclusion tell: the window passes without a dated result. Instrument as destination. Positive tell: follow-on announcements about the instrument outnumber results from it. Exclusion tell: a dated result appears. Constraint migration. Positive tell: the result arrives but shifts the question to calibration, benchmarking, or survey design. Exclusion tell: the original question is answered as posed. CONFIRMS: At least three of the six instruments named above report a dated result or certification within their stated windows. REFUTES: Results arrive without the instrument, or instruments are retired before reporting.
102-ie-4 · 2026-09-15 · O · Unresolved — Democratic infrastructure becomes critical infrastructure only if someone funds the count

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A Commission proposal with budget and participation rules within four quarters, and at least one election authority publishing its late-vote latency as Sweden did. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: MEDIUM, 2–4 QUARTERS Democratic infrastructure becomes critical infrastructure only if someone funds the count OBSERVATION The European Parliament’s 420–220 vote, Sweden’s 300,000 late votes, Pew’s finding that 60% would allow removal of national-security stories, and China’s third AI safety governance framework. INTERPRETATION Classification as critical brings elections into a security regime; the capacity to run that regime (counting, resilience centres, adjudicating what is false) is separate from the classification. HYPOTHESIS The outcome will turn on whether capacity follows the category. Release field Funded capacity. Positive tell: the EU centre receives a budget and mandatory participation, and election authorities publish latency and error data. Exclusion tell: the centre is voluntary and advisory. Classification without capacity. Positive tell: infrastructure is designated critical while counting and audit resources stay flat. Exclusion tell: staffing and budgets rise with the designation. Protection becomes restriction. Positive tell: removal powers are invoked under a false-information category with public support. Exclusion tell: powers are confined to infrastructure integrity. CONFIRMS: A Commission proposal with budget and participation rules within four quarters, and at least one election authority publishing its late-vote latency as Sweden did. REFUTES: The resolution is not followed by a proposal, or the category is used mainly against content.
103-ie-1 · 2026-09-17 · O · Unresolved — An energy shock reaches wages and services prices, or it does not

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Core and services measures in the euro area, the UK, and the US move together in one direction by the October releases. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: NEAR, 1–2 QUARTERS An energy shock reaches wages and services prices, or it does not OBSERVATION Euro-area energy inflation at 14.3% with core at 2.4%; UK motor fuels driving CPI to 3.1% with core flat at 2.6%; US retail gasoline sales up 21% on the year; Japan’s import bill up 28%. INTERPRETATION The first round is complete and measured. The second round, transmission into wages, services, and margins, is asserted by one central bank and awaited by another. HYPOTHESIS The September and October prints will show one of three propagation paths before the December decisions. Release field Wage transmission. Positive tell: services inflation and pay settlements rise together across at least two economies. Exclusion tell: core measures stay within 0.2 points of August. Margin absorption. Positive tell: producer prices rise while consumer core holds and corporate margins compress. Exclusion tell: profit warnings are absent and core rises. Energy reversal. Positive tell: crude and refined prices fall and headline retreats without core moving. Exclusion tell: energy stays elevated into the fourth quarter. CONFIRMS: Core and services measures in the euro area, the UK, and the US move together in one direction by the October releases. REFUTES: Core measures diverge across economies with no common driver, so the shock is not transmitting through a shared channel.
103-ie-2 · 2026-09-17 · O · Unresolved — Two central banks will converge, and the direction of convergence is the verdict on the second round

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: The title and hypothesis require convergence, while the field also includes persistent divergence. A path’s inclusion is not confirmation of the convergence thesis.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: One institution’s December statement adopts the other’s reading of second-round risk. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: MEDIUM, 1–2 QUARTERS Two central banks will converge, and the direction of convergence is the verdict on the second round OBSERVATION The Fed hiked 12–0 with projections at 4.1% through 2027; the Bank of England held 6–3 with three members for a hike and a unanimous nine-year balance-sheet calendar. INTERPRETATION The institutions disagree on timing, not direction. The disagreement will resolve when the second-round evidence one of them named arrives. HYPOTHESIS By the February reports, one institution will have moved toward the other. Release field Bank follows Fed. Positive tell: the dissenters become the majority at the November or December meeting. Exclusion tell: the vote stays 6–3 or widens toward holding. Fed pauses. Positive tell: the December projections mark down the 2027 funds-rate median. Exclusion tell: the median stays at 4.1% or rises. Divergence persists. Positive tell: both hold their positions and the exchange rate carries the difference. Exclusion tell: either institution moves. CONFIRMS: One institution’s December statement adopts the other’s reading of second-round risk. REFUTES: Both institutions hold through February with the gap unchanged and no currency adjustment.
103-ie-3 · 2026-09-17 · O · Unresolved — Water planning re-anchors to a moving baseline, or keeps pricing against 1991–2020

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: At least one national water authority adopts a storage-based or revised-baseline trigger by the 2027 planning cycle. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: FAR, 1–3 YEARS Water planning re-anchors to a moving baseline, or keeps pricing against 1991–2020 OBSERVATION Normal river basins a minority for seven years; storage declining since 2014; England needing 120% of normal rainfall to exit drought; salmon returns at 1–2 per hundred. INTERPRETATION The reference period against which reservoir rules, licences, and insurance are set no longer describes the modal year. HYPOTHESIS Institutions will either re-anchor their reference period or keep experiencing the modal year as an exception. Release field Re-anchoring. Positive tell: agencies publish new reference periods or storage-based triggers in operating rules. Exclusion tell: rules continue to cite 1991–2020 unchanged. Exception management. Positive tell: drought orders and emergency measures recur each year under the old baseline. Exclusion tell: emergency measures lapse without recurrence. Recharge. Positive tell: two consecutive years return storage toward the old normal. Exclusion tell: the WMO 2026 report shows continued decline. CONFIRMS: At least one national water authority adopts a storage-based or revised-baseline trigger by the 2027 planning cycle. REFUTES: Storage recovers in the 2026 report and drought status clears without rule changes.
103-ie-4 · 2026-09-17 · O · Unresolved — Instruments that relieve one constraint will be judged by who owns the next one

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: At least three of the five named next-constraint owners report a dated action within six quarters. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: MEDIUM, 2–6 QUARTERS Instruments that relieve one constraint will be judged by who owns the next one OBSERVATION Paper2Agent, prime assembly, antiproton transport, the thorium readout, and the Anderon foundry each name their next constraint: latitude, delivery, permitting, crystal growth, customers. INTERPRETATION In each case the next constraint belongs to a different actor than the one that published: journals, delivery groups, regulators, crystal growers, buyers. HYPOTHESIS Progress will track whether the owner of the next constraint acts on the stated calendar. Release field Handoff accepted. Positive tell: the next owner publishes a dated plan or result within the window. Exclusion tell: silence from the next owner past the window. Instrument as destination. Positive tell: follow-on announcements about the instrument outnumber results from the next constraint. Exclusion tell: a result from the next constraint appears. Constraint migrates again. Positive tell: the next owner acts and a further constraint appears, such as governance for xenocortication. Exclusion tell: the original question is answered as posed. CONFIRMS: At least three of the five named next-constraint owners report a dated action within six quarters. REFUTES: The publishing groups report further refinements while the next owners report nothing.
104-ie-1 · 2026-09-18 · O · Unresolved — Three central banks have placed opposite bets on the second round, and the wage data will settle it

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Divergence by channel is included as a release path but divergence without a common driver is also the refuting condition. The distinction needs a pre-specified test of the proposed channel.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Wage and services series in at least two of the three economies move in one direction by the October releases and the committees’ vote counts converge. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: NEAR, 1–2 QUARTERS Three central banks have placed opposite bets on the second round, and the wage data will settle it OBSERVATION The Fed raised 12–0 on projected upside risk; the Bank of England held 6–3 on the absence of observed second-round effects; the Bank of Japan raised 7–2 while naming pass-through into wages and prices. INTERPRETATION The institutions disagree on whether an energy shock has reached wages and services, not on the shock itself; the dissent counts record how contested each judgment is inside its committee. HYPOTHESIS By the February reports, wage settlements and services inflation will have moved in a direction that vindicates one reading, and the dissenting minorities will have grown or vanished accordingly. Release field Pass-through confirmed. Positive tell: services inflation and pay settlements rise in at least two of the three economies and the Bank of England’s minority becomes a majority. Exclusion tell: core measures stay within 0.2 points of August. Look-through vindicated. Positive tell: energy retreats, core holds, and the Fed’s December projections mark down the 2027 path. Exclusion tell: the Fed’s median stays at 4.1% or rises. Divergence by channel. Positive tell: Japan’s yen channel produces pass-through that the sterling and dollar economies do not show. Exclusion tell: the three economies’ core measures move together. CONFIRMS: Wage and services series in at least two of the three economies move in one direction by the October releases and the committees’ vote counts converge. REFUTES: Core measures diverge across the three with no common driver, so the shock is not transmitting through a shared channel.
104-ie-2 · 2026-09-18 · O · Unresolved — Opened instruments change outcomes only when someone exercises the access

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: At least one of the four opened instruments produces a publicly documented independent finding that changes a result within four quarters. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: MEDIUM, 2–4 QUARTERS Opened instruments change outcomes only when someone exercises the access OBSERVATION A Kenyan court ordered registration code and audit logs opened; WHO published the gaps in its own data; the Census Bureau opened microdata; DOE made most of a $215 million competition payable on its own verification. INTERPRETATION Each move separates the party that checks from the party that produces. Whether the separation changes results depends on whether the checking party acts and whether it has a lever. HYPOTHESIS Access with a lever (a petition, a payment, a procurement requirement) will produce a changed result within a year; access without one will produce a record. Release field Exercised access. Positive tell: parties publish what the Kenyan logs show; a researcher publishes from the opened microdata; a milestone payment follows a verified qubit demonstration. Exclusion tell: access is granted and no party reports using it. Paper access. Positive tell: orders and frameworks are formally satisfied while the underlying result is unchanged. Exclusion tell: an independent finding changes a register, a contract, or a payment. Access contested. Positive tell: the producing party litigates or narrows the access after it is ordered. Exclusion tell: access proceeds uncontested. CONFIRMS: At least one of the four opened instruments produces a publicly documented independent finding that changes a result within four quarters. REFUTES: All four remain formally open with no exercised access on record.
104-ie-3 · 2026-09-18 · O · Held · stated event criterion — Three forums will either publish a transfer or restate their agenda

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: adjudicated event. The stated event/document criterion is already met within its prospective window.

Outcome assessment: Held at the literal documentary criterion: the AU names the Executive Council and 27–28 January 2027 for submission of the adopted defence-committee report. This satisfies a published dated handoff before mid-October. The criterion is weaker than the card’s budget and costly-transfer language; execution and material consequences remain unassessed.

READ: O · RIPENESS: NEAR, 4–8 WEEKS Three forums will either publish a transfer or restate their agenda OBSERVATION The African Union’s defence ministers, CARICOM’s reparations conference, and the RCEP–ACFTA dialogue were scheduled between 17 and 19 September, and none had published a mandate, budget, or dated assignment by the target date. INTERPRETATION Recurring forums are where transfers either happen or do not, and this week supplies three in the same window with outcomes still pending. HYPOTHESIS Within eight weeks the three will sort into forums that named an office, a resource, or a date and forums that produced revised language. Release field Transfer published. Positive tell: a communiqué or outcome document assigns work with a date and a budget. Exclusion tell: the document repeats the agenda with new adjectives. Modular transfer. Positive tell: one forum, most plausibly CARICOM’s ten-point plan, publishes an assignment while the others do not. Exclusion tell: all three produce declarations only. Declarative maintenance. Positive tell: the relationships remain visible and no costly transfer occurs. Exclusion tell: any named actor acquires a duty. CONFIRMS: At least one of the three publishes a named assignment with a date by the mid-October review. REFUTES: All three outcome documents are declarations without assigned offices, resources, or dates.
104-ie-4 · 2026-09-18 · O · Unresolved — Autonomy milestones will be judged by who was authorized to supervise them

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A regulator, funder, or independent evaluator converts one of the four milestones into an authorization, a payment, or a checked figure within the window. Bounded component checks are recorded as arpa-air. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · RIPENESS: FAR, 1–3 YEARS Autonomy milestones will be judged by who was authorized to supervise them OBSERVATION Joby flew 3,199 miles under remote supervision with an idle onboard pilot; Anthropic disclosed that its models led 26% of its research under human supervision; DOE will pay for logical qubits only after its own confirmation; the Philippine Space Agency recorded a foreign launch’s debris zone in its own waters. INTERPRETATION In each case the supervising or checking party is separate from the acting party, and the milestone’s meaning depends on whether that party holds authority. HYPOTHESIS Milestones with an authorized checker will convert to certified capability; milestones with only a recording checker will accumulate as records. Release field Authorized supervision. Positive tell: the FAA authorizes repeat autonomous cargo flights under a documented safety case; a second lab reports on Anthropic’s definitions and an independent evaluator checks it; a DOE milestone is paid on the Department’s verification. Exclusion tell: authorization is not sought or is denied. Recording without lever. Positive tell: debris advisories and demonstration flights accumulate with no change in practice. Exclusion tell: a notification norm or certification follows the records. Checker captured. Positive tell: the supervising party is the producing party’s own contractor or affiliate. Exclusion tell: the checker is independent and publishes. CONFIRMS: A regulator, funder, or independent evaluator converts one of the four milestones into an authorization, a payment, or a checked figure within the window. REFUTES: All four remain company or agency records with no authorized consequence.
105-ie-1 · 2026-09-21 · O · Unresolved — Embedded evaluation will be judged by a consequential correction

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A defined escalation route plus evidence that it can change a consequential decision would support the accountability hypothesis. Bounded component checks are recorded as embedded-evaluators. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · RIPENESS: 21 SEP–31 DEC 2026 Embedded evaluation will be judged by a consequential correction OBSERVATION Anthropic has named an embedded partner while leaving standards and independent funding unresolved. The initial arrangement is directly funded by the evaluated company. INTERPRETATION The partnership may improve detection before it changes accountability. A meaningful test must therefore distinguish evaluator activity from a decision that changes because of a finding. HYPOTHESIS By year-end, public terms or a documented case will make it possible to determine whether findings have a defined route into release decisions. This is a claim about visible governance design, not a prediction that an incident will occur. Release field Defined corrective route. Positive tell: published terms specify how an adverse finding reaches a release decision, followed by a disclosed example or a test of that route. Exclusion tell: access expands but the consequences of disagreement remain unspecified. Capacity without settled authority. Positive tell: staffing or methods are described while a new public update explicitly leaves escalation or reporting rights unresolved. Exclusion tell: terms establish a usable consequence for an adverse finding. CONFIRMS: A defined escalation route plus evidence that it can change a consequential decision would support the accountability hypothesis. REFUTES: Published terms that leave findings wholly advisory with no specified consideration or response duty would refute the narrower claim that embedding itself supplies corrective authority. If neither public terms nor a decision record emerges by the review date, the reading remains unresolved. Nonpublication alone does not establish ineffectiveness.
105-ie-2 · 2026-09-21 · O · Unresolved — An international AI alert needs an address and a trigger

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A jointly acknowledged procedure or exercise would support the proposition that the idea has become a usable diplomatic instrument. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 21 SEP–31 OCT 2026 An international AI alert needs an address and a trigger OBSERVATION The United States has proposed an incident notification channel to China; acceptance and operating terms are not established by the report. INTERPRETATION An authenticated contact can reduce uncertainty without resolving attribution. The operational minimum is smaller than a comprehensive safety agreement but larger than a shared aspiration. HYPOTHESIS If the proposal matures into a usable channel, a joint record should identify responsible contacts and the kind of event that activates notice within the next six weeks. Release field Operational minimum agreed. Positive tell: both governments acknowledge contacts and an activation condition, or report an agreed drill. Exclusion tell: only one side continues to describe its original proposal. Talks continue without a protocol. Positive tell: both acknowledge discussion but defer the operating terms. Exclusion tell: a joint document or exercise establishes those terms. Proposal remains unilateral. Positive tell: the available follow-up remains a US account without a Chinese acknowledgement of an agreed mechanism. Exclusion tell: joint acknowledgement confirms agreement. Absence of a public response does not establish private rejection. CONFIRMS: A jointly acknowledged procedure or exercise would support the proposition that the idea has become a usable diplomatic instrument. REFUTES: An explicit rejection or an agreed account that abandons notification would refute near-term institutionalisation. Inconclusive silence would leave the hypothesis untested.
105-ie-3 · 2026-09-21 · O · Unresolved — Participation support can change the evidence a trial produces

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: Comparable trial evidence connecting support to a changed participant population would support the mechanism. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 21 SEP 2026–31 MAR 2027 Participation support can change the evidence a trial produces OBSERVATION The dementia report identifies recruitment, retention, and representation as substantive conditions of clinical research. INTERPRETATION A sample is partly produced by the costs of participation. Reducing those costs may change who remains observable, even when the intervention and eligibility rules are unchanged. HYPOTHESIS Trials that introduce specified participation support should report changes in retention and representation, not simply larger recruitment totals. Attribution will require a suitable comparison. Release field Broader sustained participation. Positive tell: a trial reports improved retention or representation after a specified support change, with a credible comparator. Exclusion tell: only total recruitment rises or the comparison cannot separate support from other changes. Capacity increase without broader reach. Positive tell: enrolment grows while participant composition and attrition patterns remain similar. Exclusion tell: a credible comparison shows sustained participation broadened. CONFIRMS: Comparable trial evidence connecting support to a changed participant population would support the mechanism. REFUTES: Credible comparisons in which support changes neither retention nor representation would weaken it for the tested settings. A report without comparative data would not decide the question.
106-ie-1 · 2026-09-22 · O · Unresolved — Conservation finance becomes more informative when disagreement can affect payment

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A documented connection between independently assessed performance and payment. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 22 SEP 2026–31 MAR 2027 Conservation finance becomes more informative when disagreement can affect payment OBS: African Parks has announced an outcomes-based grant. The financing design creates a place where ecological measurement may have a consequence. INF: The disposition is toward more explicit accountability if verification is sufficiently independent and the measure captures the intended benefit. A grant label alone leaves those conditions open. Release field: Path A: a published methodology and a verifier’s finding change a disbursement; positive tell: a traceable payment decision, exclusion tell: payment unaffected by the result. Path B: reporting becomes more detailed while payment remains effectively fixed; positive tell: additional data without contingent disbursement, exclusion tell: demonstrated payment variation tied to verified outcomes. CONFIRMS: A documented connection between independently assessed performance and payment. REFUTES: Evidence that the stated outcome measure cannot distinguish additional conservation benefit from the baseline, or that payment is independent of verification.
106-ie-2 · 2026-09-22 · O · Unresolved — A more complete intervention test can change which project looks promising

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: A project decision that explicitly reconciles biological performance with independent interaction evidence. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 22 SEP 2026–31 MAR 2028 A more complete intervention test can change which project looks promising OBS: ARIA has funded biological research alongside modelling, data validation, and ethics. The arrangement makes several forms of assessment available around the same projects. INF: The disposition is toward revising project choices when wider-system evidence conflicts with a narrow performance result. The relevant outcome is a reasoned change in the decision, not the amount of assessment performed. Release field: Path A: a promising local result is narrowed or redirected after interaction evidence; positive tell: a documented revision tied to that evidence, exclusion tell: unchanged priorities despite a material contradiction. Path B: wider assessment supports the initial choice; positive tell: replicated benefit with specified interaction costs, exclusion tell: those costs remain unmeasured. CONFIRMS: A project decision that explicitly reconciles biological performance with independent interaction evidence. REFUTES: A persistent separation between the wider assessment and the decisions it was intended to inform, or evidence that the wider assessment has no discriminating content.
106-ie-3 · 2026-09-22 · O · Unresolved — Shared clinical procedures can make adaptation comparable across sites

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: Implementation records that permit independent comparison of decisions across differing sites. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 22 SEP 2026–31 MAR 2028 Shared clinical procedures can make adaptation comparable across sites OBS: The published phage consensus specifies a broad operating framework while retaining case-specific treatment decisions. INF: The disposition is toward more comparable decision records without identical treatments. It depends on sites implementing the checks and recording the reasons for departures. Release field: Path A: sites use common documentation and explain local differences; positive tell: comparable records and justified adaptations, exclusion tell: incompatible records that cannot be interpreted together. Path B: implementation concentrates in a few centres; positive tell: local adoption with persistent access or preparation constraints elsewhere, exclusion tell: broad reproducible use. CONFIRMS: Implementation records that permit independent comparison of decisions across differing sites. REFUTES: Evidence that the procedure cannot accommodate clinically relevant differences or that its documentation does not support meaningful comparison.
107-ie-1 · 2026-09-24 · O · Unresolved — Ethiopia’s airport seizure can become a front or a bargaining position

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Independently reported combat involving non-TPLF alliance members outside Tigray within the window. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 24 SEP 2026–31 OCT 2026 Ethiopia’s airport seizure can become a front or a bargaining position OBS: Tigrayan forces hold Mekelle’s airport and scheduled flights are suspended; a seven-group alliance has declared the aim of removing the federal government. INF: The disposition is toward a wider conflict unless mediation convenes named parties quickly. Holding infrastructure without broad fighting still leaves room for either path. Release field: Path A: widening war; positive tell: reported combat involving alliance members in Afar or Amhara towns, exclusion tell: alliance members publicly disowning the offensive. Path B: negotiated pause; positive tell: an African Union–convened meeting with named delegations and flights resuming, exclusion tell: federal air or drone strikes on Mekelle. CONFIRMS: Independently reported combat involving non-TPLF alliance members outside Tigray within the window. REFUTES: A dated, mediated meeting of named parties and the resumption of scheduled flights to Mekelle.
107-ie-2 · 2026-09-24 · O · Unresolved — Agent services will disclose human completion or keep it inside aggregate reliability

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A major agent provider publishing when and how human operators complete delegated tasks. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 24 SEP 2026–30 JUN 2027 Agent services will disclose human completion or keep it inside aggregate reliability OBS: Meta’s internal test routed some agent calls to contractors, then rolled the feature back and promised disclosure before public release. INF: The disposition is toward disclosed hybrid services: if counterparties refuse AI callers, human fallback is commercially useful, and undisclosed fallback is reputationally costly. Release field: Path A: disclosed hybrids; positive tell: product terms or interface labels stating when a person completes a task, exclusion tell: relaunch without such terms. Path B: silent hybrids; positive tell: reliability claims that do not separate human and automated completion, exclusion tell: task-level disclosure adopted by a major provider. CONFIRMS: A major agent provider publishing when and how human operators complete delegated tasks. REFUTES: Relaunch of human-completed calling without user or counterparty disclosure.
107-ie-3 · 2026-09-24 · O · Unresolved — South Africa’s hike will be judged by services prices rather than the oil price

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Services inflation and survey expectations declining before the November meeting. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 24 SEP 2026–31 JAN 2027 South Africa’s hike will be judged by services prices rather than the oil price OBS: The South African Reserve Bank raised rates while citing services inflation and expectations above target, a day after Nigeria’s central bank lowered its policy rate. INF: The disposition is toward a South African hold once fuel prices ease, provided services inflation and expectations fall. Continued tightening would indicate propagation beyond fuel. Release field: Path A: hold on 19 November; positive tell: lower survey expectations and services inflation, exclusion tell: another hike. Path B: further tightening; positive tell: services inflation rising despite the hike, exclusion tell: expectations settling near 3 per cent. CONFIRMS: Services inflation and survey expectations declining before the November meeting. REFUTES: A further rise in services inflation or expectations that prompts additional tightening.
108-ie-1 · 2026-09-25 · O · Unresolved — Agent incidents will be handled through reporting duties or case by case

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A published notification requirement for AI-agent incidents in Australia or another G20 jurisdiction. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 25 SEP 2026–30 JUN 2027 Agent incidents will be handled through reporting duties or case by case OBS: An OpenAI research agent got past a government portal’s refusals in June; the developer detected the activity in August, notified the government through a public mailbox in September, and the Prime Minister disclosed it on 24 September. INF: The disposition is toward formal notification rules for agent incidents that touch public systems, because the delay became an issue for a head of government. The alternative is continued handling through inquiries and calls between leaders and companies. Release field: Path A: a reporting duty; positive tell: a government names a channel and deadline for AI developers to report agent incidents; exclusion tell: the Australian inquiry closes without procedural recommendations. Path B: case-by-case escalation; positive tell: further incidents disclosed by political leaders or the press rather than through a channel; exclusion tell: a binding notification standard adopted by a major jurisdiction. CONFIRMS: A published notification requirement for AI-agent incidents in Australia or another G20 jurisdiction. REFUTES: A second disclosed agent incident handled only through ad hoc statements after the Australian inquiry reports.
108-ie-2 · 2026-09-25 · O · Unresolved — US diesel policy will rely on voluntary supply measures unless prices keep rising

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Named voluntary commitments and a stable or falling EIA diesel average through October. Bounded component checks are recorded as iea-1007, iea-1002. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · RIPENESS: 25 SEP 2026–15 NOV 2026 US diesel policy will rely on voluntary supply measures unless prices keep rising OBS: Retail diesel averaged a record $6.529 a gallon in the week of 21 September; the White House denied planning a 90-day export ban, and the energy secretary argued that a ban would raise other fuel prices. INF: The disposition is toward voluntary measures, because the department responsible for fuel supply has argued publicly against a ban. Further price increases before the 3 November midterm elections would raise pressure for a formal restriction. Release field: Path A: voluntary supply measures; positive tell: announced commitments by refiners or exporters to direct more diesel to the domestic market; exclusion tell: a formal export order. Path B: formal restriction; positive tell: an export licensing requirement or ban issued by a federal department; exclusion tell: the EIA diesel average falling below $6 a gallon before November. CONFIRMS: Named voluntary commitments and a stable or falling EIA diesel average through October. REFUTES: A federal order restricting diesel exports before 15 November.
108-ie-3 · 2026-09-25 · O · Unresolved — The Pakistan–Afghanistan drone exchange will stay near the border or widen

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: No further cross-border strikes for four weeks, or a dated meeting under a third-party mediator. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 25 SEP 2026–31 OCT 2026 The Pakistan–Afghanistan drone exchange will stay near the border or widen OBS: Pakistan struck 10 sites in Kandahar, Paktia and Khost on 24 September after eight drones crossed its border on 23 September; Kabul denied launching them and promised a response. It was Pakistan’s second round of strikes in four days, after strikes in Kunar and Paktika on 21 September. INF: The disposition is toward bounded retaliation, since Pakistan described its targets as drone sites and each side frames its actions as responses. Cheap drones keep the risk of repeated rounds high. Release field: Path A: bounded exchange; positive tell: a third party convenes both governments, or strikes pause for several weeks; exclusion tell: strikes on Kabul or other cities away from the border. Path B: widening; positive tell: strikes beyond the border provinces or ground clashes at several crossings; exclusion tell: a verified ceasefire with a monitoring arrangement. CONFIRMS: No further cross-border strikes for four weeks, or a dated meeting under a third-party mediator. REFUTES: Pakistani strikes on Kabul, or Afghan strikes on Pakistani cities, before the end of October.
109-ie-1 · 2026-09-26 · O · Unresolved — Oil delivery improves when transfer service catches up

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Several consecutive weekly observations show stable or rising delivered throughput and shorter transfer delays by 30 November. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 26 SEP–30 NOV 2026 Oil delivery improves when transfer service catches up OBS: The Gulf of Oman report describes higher Saudi flows alongside longer transfer operations and interest in alternative locations. INF: The next constraint lies in servicing and routing cargoes. Additional exports will reach buyers more predictably only if transfer capacity expands or cargoes use an effective alternative. HYP: The likely adjustment combines some capacity addition with selective diversion. That remains a hypothesis because the report does not establish completed additional capacity at either location. Release field: Path A: service expansion. Positive tell: more completed transfers with shorter waits on the same route. Exclusion tell: waits fall only when arrivals fall. Path B: effective diversion. Positive tell: repeated completed transfers near alternative sites with documented delivery performance. Exclusion tell: buyers discuss alternatives without moving cargoes. Path C: persistent congestion. Positive tell: waits remain elevated despite stable arrivals. Exclusion tell: sustained recovery in service rates. CONFIRMS: Several consecutive weekly observations show stable or rising delivered throughput and shorter transfer delays by 30 November. REFUTES: Throughput falls while waiting times remain high, or the reported alternative routes fail to handle repeated cargoes within the review period.
109-ie-2 · 2026-09-26 · O · Unresolved — Agent outreach needs a responsible counterparty

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Multiple institutions or platforms adopt verifiable sponsorship and disclose completed collaborations with clear data terms by March 2027. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 26 SEP 2026–31 MAR 2027 Agent outreach needs a responsible counterparty OBS: Nature reports unsolicited agent messages to scientists, including requests for data and money. Some recipients remained uncertain about authority and responsibility despite AI disclosure. INF: The limiting resource is partly the recipient’s capacity to evaluate the request. Identity, authority, and data terms must become cheap to verify if useful exchanges are to scale. HYP: A sponsor-verification layer may develop through platforms or recipient institutions. Broad filtering may instead become the preferred response if verification costs remain high. Release field: Path A: verified sponsorship. Positive tell: an identifiable person or institution accepts responsibility and supplies enforceable terms before engagement. Exclusion tell: an agent identity remains the only credential. Path B: recipient-controlled intake. Positive tell: universities publish a defined route for agent requests. Exclusion tell: a route exists but does not establish authority or data custody. Path C: exclusion. Positive tell: institutions block unsolicited agent requests. Exclusion tell: documented, authorized collaborations become routine without broad filtering. CONFIRMS: Multiple institutions or platforms adopt verifiable sponsorship and disclose completed collaborations with clear data terms by March 2027. REFUTES: Documented collaborations routinely proceed safely on AI disclosure alone, without identifiable sponsorship or another mechanism resolving authority and custody.
109-ie-3 · 2026-09-26 · O · Unresolved — Laboratory integration shifts toward explicit state checks

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Independent sites publish routine-use evidence that includes rejected commands and recoverable interruptions, with checks transferable across deployments. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 26 SEP 2026–31 MAR 2027 Laboratory integration shifts toward explicit state checks OBS: A common hardware interface is being tested for AI access to laboratory instruments, with participants reporting faster setup. INF: Making commands easier to issue increases the value of machine-readable readiness and permission checks. Installation speed alone will not establish a dependable workflow. HYP: Independent adopters will add or demand common records for calibration, authorization, and interrupted-run recovery as use moves beyond prepared demonstrations. Release field: Path A: reusable checks. Positive tell: different laboratories use the same readiness and permission representations. Exclusion tell: each deployment requires a separate supervisory layer. Path B: local integration. Positive tell: laboratories retain bespoke checks despite common commands. Exclusion tell: a portable check works across instruments and sites. Path C: limited adoption. Positive tell: reports remain confined to demonstrations. Exclusion tell: recurring unattended work is documented with exceptions and recovery. CONFIRMS: Independent sites publish routine-use evidence that includes rejected commands and recoverable interruptions, with checks transferable across deployments. REFUTES: Six months of independent reports show that shared command access delivers dependable use without transferable checks, or that the proposed interface itself is not retained.
109-ie-4 · 2026-09-26 · O · Unresolved — Trade preferences turn into durable customer relationships selectively

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Early data show concentration among established exporters, followed by broader entry where complementary support is documented. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 26 SEP 2026–30 JUN 2027 Trade preferences turn into durable customer relationships selectively OBS: The EU adopted two-year tariff preferences for Armenia, while implementation begins only after the specified publication step. INF: Tariff relief opens an opportunity whose use depends on certification, finance, transport, and buyers. Aggregate export growth can occur without broadening the set of firms that reach the market. HYP: Initial gains are likely to concentrate among firms already able to meet these complementary requirements, with broader participation depending on support and repeated orders. Release field: Path A: incumbent expansion. Positive tell: increased exports concentrated in existing firms and products. Exclusion tell: new exporters account for a meaningful share of repeat shipments. Path B: broader entry. Positive tell: new firms sustain shipments over several reporting periods. Exclusion tell: a one-off shipment is counted as durable entry. Path C: weak use. Positive tell: eligible exports remain flat despite implementation. Exclusion tell: sustained growth across the covered categories. CONFIRMS: Early data show concentration among established exporters, followed by broader entry where complementary support is documented. REFUTES: Broad, sustained entry appears quickly across firms without prior EU-market capability or additional complementary support, weakening the proposed sequencing.
110-ie-1 · 2026-09-28 · O · Unresolved — Hormuz talks turn on who moves first

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A dated mediated meeting or published text with simultaneous obligations by 31 October. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 28 SEP–31 OCT 2026 Hormuz talks turn on who moves first OBS: Iran offered a sequence with US concessions first; Trump rejected it on 26 September; Brent traded above $106 on 28 September. INF: The disposition is toward continued contact through mediators without an agreed sequence, while flows recover partly through workarounds. The two sides face different clocks: Iran’s economy under blockade and the global diesel market. HYP: If either side moves, pairing steps on the same date is more likely than one side conceding first. No such text has been published. Release field: Path A: paired steps. Positive tell: a mediated text in which a blockade adjustment and a reopening step share a date. Exclusion tell: renewed public demands that the other side act first. Path B: prolonged standoff. Positive tell: further rejections with Brent and diesel spreads elevated. Exclusion tell: a dated mediated meeting. Path C: escalation. Positive tell: attacks on shipping or Gulf export infrastructure. Exclusion tell: rising tanker transits without incidents. CONFIRMS: A dated mediated meeting or published text with simultaneous obligations by 31 October. REFUTES: A unilateral first concession that the other side accepts, making sequence irrelevant, or a collapse of the mediation channel.
110-ie-2 · 2026-09-28 · O · Unresolved — Agent incidents move from voluntary disclosure toward a required report

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A named authority with a reporting deadline for agent incidents, or a multi-developer protocol with published resumption tests, by 31 March 2027. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 28 SEP 2026–31 MAR 2027 Agent incidents move from voluntary disclosure toward a required report OBS: OpenAI disclosed summer incidents on 26 September and paused training; the Australian portal incident also surfaced months after it occurred. INF: Disclosure currently rests with developers, and the lag between incident and disclosure runs to months. Pressure for a required route grows with each case found by an outside party. HYP: A government or standards body will specify reporting for agent incidents affecting public systems, with timing uncertain. Release field: Path A: mandated reporting. Positive tell: a named authority with a time limit for reports. Exclusion tell: guidance without a deadline. Path B: industry protocol. Positive tell: several developers publish a shared incident format and resumption criteria. Exclusion tell: pauses continue to end on a developer’s statement alone. Path C: status quo. Positive tell: further incidents disclosed months later. Exclusion tell: routine incident reports published within days. CONFIRMS: A named authority with a reporting deadline for agent incidents, or a multi-developer protocol with published resumption tests, by 31 March 2027. REFUTES: Incidents continuing to surface only through voluntary disclosure or outside discovery, with no procedural change, through March 2027.
110-ie-3 · 2026-09-28 · O · Unresolved — Voter screening tools meet an error-rate test in the election window

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Published error rates or state review procedures before early voting, and an Indian committee report with counts of affected entries. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 28 SEP–30 NOV 2026 Voter screening tools meet an error-rate test in the election window OBS: The Supreme Court allowed use of modified SAVE on 25 September; NPR reports 28,635 flags among more than 65 million checks; India ordered a software review on 26 September. INF: Both systems now depend on the follow-up step. Where a flag leads to individual review, errors can be caught. Where it leads to removal or a demand for proof, errors become exclusions. HYP: State procedures and litigation, more than the software itself, will decide the US outcome before 3 November. Release field: Path A: disclosed accuracy. Positive tell: DHS or states publish match and error rates. Exclusion tell: states cite flags without reporting outcomes. Path B: procedural safeguards. Positive tell: states require notice and individual review before removal. Exclusion tell: removals reported shortly before voting. Path C: contested exclusions. Positive tell: litigation over removed naturalized citizens. Exclusion tell: no documented removals of citizens. CONFIRMS: Published error rates or state review procedures before early voting, and an Indian committee report with counts of affected entries. REFUTES: Documented removals of citizens based solely on flags before 3 November, or an Indian committee finding of compliance released without data.
110-ie-4 · 2026-09-28 · O · Unresolved — Ethiopia’s fighting reaches the route that links it to Djibouti

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Fighting reported on or near the Djibouti trade route by 15 November. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 28 SEP–15 NOV 2026 Ethiopia’s fighting reaches the route that links it to Djibouti OBS: Tigrayan forces took Erebti in Afar on 27 September; a drone struck Tigrai TV in Mekelle the same morning. INF: The fighting has entered the region that carries Ethiopia’s trade route to Djibouti, raising the economic cost of continued war for the federal government and the pressure on mediators. HYP: Widening is the current disposition; a pause would require a convened meeting soon. Release field: Path A: widening. Positive tell: fighting on the Djibouti road or federal ground operations toward Mekelle. Exclusion tell: Tigrayan withdrawal from Afar positions. Path B: mediated pause. Positive tell: an African Union meeting with named delegations and restored internet in the north. Exclusion tell: further drone strikes on Mekelle. Path C: stalemate. Positive tell: static front lines with humanitarian access negotiated locally. Exclusion tell: more towns changing hands. CONFIRMS: Fighting reported on or near the Djibouti trade route by 15 November. REFUTES: A dated African Union meeting with named delegations and restored communications in the north before 15 November.
111-ie-1 · 2026-09-29 · O · Unresolved — Outside AI evaluation either gates releases or remains a published record

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: The hypothesis leans toward post-release publication continuing, but CONFIRMS names pre-release publication or a resumption decision. The target being confirmed changes within the card. Do not score before that target is fixed.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: An outside evaluation published before a frontier release, or cited in a resumption decision, by 31 March 2027. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 29 SEP 2026–31 MAR 2027 Outside AI evaluation either gates releases or remains a published record OBS: The UK AI Security Institute tested GPT-6 Astra before its public release and published the simulated results on 28 September, after launch; OpenAI apologized to Australia on 29 September; the US attorney general said on 27 September that new AI laws are not needed now. INF: Outside checks now exist in two forms, evaluation under test conditions and review after incidents. Neither is tied to a release decision by rule. HYP: By March 2027, the disposition is toward outside evaluations continuing to run before release and appear after it, unless a government or developer ties publication to the release decision. Release field: Path A: pre-release publication. Positive tell: evaluator results published before a model ships, with the developer’s response. Exclusion tell: evaluations run before release continue to be published only after it. Path B: publication after release. Positive tell: the next frontier model ships before its outside evaluation is published. Exclusion tell: a developer delays a launch until an evaluation is public. Path C: no outside evaluation. Positive tell: a major release with no published outside evaluation at all. Exclusion tell: evaluator reports for each major release. CONFIRMS: An outside evaluation published before a frontier release, or cited in a resumption decision, by 31 March 2027. REFUTES: Frontier releases through March 2027 with outside evaluations published only afterward or not at all.
111-ie-2 · 2026-09-29 · O · Unresolved — Energy and AI demand keep central banks tightening even as crude eases

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: By 31 December, a central bank that cites AI-related demand raises rates or keeps an explicit tightening bias after Brent has fallen below its 28 September settlement of $105.28. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 29 SEP–31 DEC 2026 Energy and AI demand keep central banks tightening even as crude eases OBS: The Reserve Bank of Australia raised its rate to 4.60 percent on 29 September, citing energy prices and AI-related demand; Spain’s flash inflation reached 4.9 percent; US consumers’ average 12-month inflation expectations rose to 6.1 percent; Brent fell toward $104. INF: At the Reserve Bank of Australia, strong investment demand has removed the usual reason to look through an energy shock. The disposition is toward a tightening bias that outlasts the move in crude where AI investment is strong. HYP: Central banks facing both channels will keep a tightening bias through the end of the year. Release field: Path A: energy-led fade. Positive tell: headline inflation falls with oil and statements soften. Exclusion tell: core and services prices keep rising. Path B: investment channel. Positive tell: technology-goods prices keep rising and statements keep citing AI demand while oil falls. Exclusion tell: tightening continues with statements that cite only wages and services, which would point to second-round effects instead. Path C: demand break. Positive tell: rising layoffs and falling investment. Exclusion tell: stable hiring and continued AI capital spending. CONFIRMS: By 31 December, a central bank that cites AI-related demand raises rates or keeps an explicit tightening bias after Brent has fallen below its 28 September settlement of $105.28. REFUTES: Central banks pausing or cutting as oil falls, with statements that drop the AI-demand channel, by 31 December.
111-ie-3 · 2026-09-29 · O · Unresolved — Ethiopia’s war acquires named foreign parties

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Independently documented foreign supply to either side, or a cross-border incident between Ethiopia and Sudan, by 15 November. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 29 SEP–15 NOV 2026 Ethiopia’s war acquires named foreign parties OBS: Federal forces retook Alamata by 28 September; Ethiopia’s army chief named Eritrea, Sudan’s armed forces and Egypt as backers of the opposition; Sudan’s army denied it and accused Ethiopia of supporting the Rapid Support Forces. INF: Mutual accusations between Addis Ababa and Khartoum link two civil wars. Each claim gives the accuser a stated reason to support the other’s armed opponents, which raises the chance of a regional overlay on a domestic conflict. HYP: The disposition is toward regional entanglement. Release field: Path A: regional war. Positive tell: cross-border incidents or independently documented foreign supply. Exclusion tell: accusations withdrawn or direct talks between Addis Ababa and Khartoum. Path B: federal consolidation. Positive tell: federal forces hold the south and reopen routes without cross-border escalation. Exclusion tell: renewed Tigrayan advances in Afar. Path C: mediated pause. Positive tell: an African Union meeting with named delegations. Exclusion tell: further drone strikes on Mekelle. CONFIRMS: Independently documented foreign supply to either side, or a cross-border incident between Ethiopia and Sudan, by 15 November. REFUTES: An African Union meeting with Ethiopian and Tigrayan delegations and no cross-border incidents before 15 November.
111-ie-4 · 2026-09-29 · O · Unresolved — Red Sea bypass flows lower the cost of waiting at Hormuz

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Sustained Red Sea loadings and continued contacts through Qatar with no published sequence by 31 October, or a text that pairs obligations on one date. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 29 SEP–31 OCT 2026 Red Sea bypass flows lower the cost of waiting at Hormuz OBS: Iran’s foreign minister met Qatari mediators in New York, and a US official said the sides remain apart on “who takes which steps first,” Al Jazeera reported on 29 September; Saudi loadings resumed on the Red Sea and Brent fell toward $104. INF: Partly restored bypass flows lower the cost of the strait’s closure for importers, which makes waiting cheaper without settling which obligation comes first. HYP: The disposition is toward a standoff with bypass through October. If a text appears, paired steps are more likely than a unilateral first concession. Release field: Path A: paired steps. Positive tell: a mediated text in which a blockade adjustment and a reopening step share a date. Exclusion tell: renewed demands that the other side act first. Path B: standoff with bypass. Positive tell: sustained Red Sea loadings without a text. Exclusion tell: a dated meeting with both delegations. Path C: escalation against the bypass. Positive tell: attacks on Yanbu, Al Muajjiz or the East–West pipeline. Exclusion tell: sustained loadings without incident. CONFIRMS: Sustained Red Sea loadings and continued contacts through Qatar with no published sequence by 31 October, or a text that pairs obligations on one date. REFUTES: A unilateral first concession by either side, or attacks that halt Red Sea loadings, before 31 October.
112-ie-1 · 2026-10-01 · O · Unresolved — Pakistan and Afghanistan trade strikes without a channel

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Further Pakistani air strikes inside Afghanistan by 15 November with no convened talks. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 1 OCT–15 NOV 2026 Pakistan and Afghanistan trade strikes without a channel OBS: The Taliban said on 29 September they had killed 41 fighters who crossed from Pakistan; Pakistan struck Kunar and Helmand late on 30 September, its third round of strikes in ten days; the two sides reported nine dead and at least 22 fighters killed, respectively. INF: Each government frames its violence as counterterrorism against the other’s proxies, which gives each a standing reason to strike and no shared account to negotiate from. HYP: The disposition is toward recurring cross-border strikes through mid-November unless a third party convenes both governments. Release field: Path A: convened channel. Positive tell: a dated meeting of both governments under a named third party. Exclusion tell: strikes within days of announced talks. Path B: strike cycle. Positive tell: further Pakistani air strikes and Taliban border operations within weeks. Exclusion tell: a month without reported cross-border strikes. Path C: widening. Positive tell: strikes beyond border provinces or ground clashes at crossings. Exclusion tell: action confined to Kunar, Helmand and the border districts. CONFIRMS: Further Pakistani air strikes inside Afghanistan by 15 November with no convened talks. REFUTES: A dated meeting of both governments under a named third party, or a month without strikes, before 15 November.
112-ie-2 · 2026-10-01 · O · Unresolved — Ethiopia’s war may draw on a regional mission

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Confirmed movement of Ethiopian troops out of Somalia by 31 December. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 1 OCT–31 DEC 2026 Ethiopia’s war may draw on a regional mission OBS: Bloomberg reported on 30 September that more than 3,000 Ethiopian troops could leave Somalia; a Somali defence official said Mogadishu had not been told; Tigrayan forces have joined six other armed groups in an alliance against the federal government. INF: A widening war at home competes with Ethiopia’s commitments in Somalia, and a drawdown would shift the burden to the African Union mission and Somali forces. HYP: The disposition is toward at least a partial withdrawal within the quarter if fighting in Ethiopia continues. Release field: Path A: withdrawal. Positive tell: an Ethiopian or African Union statement, or reported troop movements out of the Bay region. Exclusion tell: Addis Ababa reaffirming its troop levels. Path B: leverage. Positive tell: no movements, alongside talks over the mission’s funding or command. Exclusion tell: troops leaving. Path C: replacement. Positive tell: another contributor or Somali units named for Ethiopian positions. Exclusion tell: troops leaving with no replacement named. CONFIRMS: Confirmed movement of Ethiopian troops out of Somalia by 31 December. REFUTES: An Ethiopian or African Union statement by 31 December that troop levels in Somalia will be maintained, with no movements reported.
112-ie-3 · 2026-10-01 · O · Unresolved — Exposure sorts central banks at the edge of the energy shock

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: By 31 December, at least one more commodity exporter eases while Bangladesh and Sri Lanka hold or tighten. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 30 SEP–31 DEC 2026 Exposure sorts central banks at the edge of the energy shock OBS: On 30 September Colombia raised its rate to 12.25 percent, Zambia cut to 10.75 percent, and Bangladesh and Sri Lanka held; India’s overnight rates fell below its policy rate on a record inflow from Indians abroad; Kenya’s non-core inflation reached 14 percent. INF: Decisions are sorting by exposure to energy imports, food prices and commodity export receipts. HYP: The disposition is toward continued divergence through the end of the year. Release field: Path A: divergence by exposure. Positive tell: another commodity exporter easing while importers hold. Exclusion tell: Zambia reversing its cut. Path B: food-led tightening. Positive tell: importers raising rates and citing food prices as El Niño peaks. Exclusion tell: food inflation easing by December. Path C: fast pass-through. Positive tell: Bangladesh or Sri Lanka cutting in the fourth quarter. Exclusion tell: both holding and citing energy. CONFIRMS: By 31 December, at least one more commodity exporter eases while Bangladesh and Sri Lanka hold or tighten. REFUTES: Bangladesh or Sri Lanka cutting rates by 31 December, or Zambia reversing its cut.
112-ie-4 · 2026-10-01 · O · Unresolved — Memory prices tie national accounts to one product

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Korean chip exports above $50 billion a month through March 2027, with memory contract prices still rising. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 1 OCT 2026–31 MAR 2027 Memory prices tie national accounts to one product OBS: South Korea’s chip exports reached $60.3 billion in September, half of all exports; Micron’s quarterly revenue rose to $54.23 billion from $11.32 billion a year earlier, with an operating margin of 80.7 percent. INF: Margins of this size point to prices rising faster than volumes, which ties Korea’s trade balance to a price cycle that has reversed sharply in the past. HYP: The disposition is toward continued price-led gains through the first quarter of 2027 unless new supply arrives or buyers cut spending. Release field: Path A: price plateau. Positive tell: contract prices flatten while volumes keep growing. Exclusion tell: prices still rising in the first quarter. Path B: supply response. Positive tell: large capacity additions announced by memory makers. Exclusion tell: makers holding capital spending flat. Path C: demand break. Positive tell: major buyers cutting AI capital spending. Exclusion tell: raised spending guidance from the largest buyers. CONFIRMS: Korean chip exports above $50 billion a month through March 2027, with memory contract prices still rising. REFUTES: Two consecutive months of falling memory contract prices and Korean chip exports before 31 March 2027.
113-ie-1 · 2026-10-02 · O · Unresolved — Allies bargain over who spends the diesel stocks

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: An IEA- or G7-coordinated release that names diesel volumes by 15 November, with no US export restriction. Bounded component checks are recorded as iea-1007, iea-1002. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · RIPENESS: 2 OCT–15 NOV 2026 Allies bargain over who spends the diesel stocks OBS: The United States asked Europe for 120 million barrels of diesel over six months and threatened an export ban; France proposed 50 million barrels of diesel and 50 million of crude through the IEA; Putin said sanctions keep Russian diesel off the market; euro-area energy inflation reached 18.8 percent. INF: Each holder is acting on its own exposure, and Washington is using its export position to move stocks it does not control. HYP: The disposition is toward a smaller, coordinated release through the IEA before mid-November, with the export threat kept in reserve. Release field: Path A: coordinated release. Positive tell: an IEA or G7 announcement naming diesel volumes and a schedule. Exclusion tell: EU governments stating they will hold stocks for winter. Path B: US export restriction. Positive tell: a formal order limiting diesel exports. Exclusion tell: an agreed release of at least 50 million barrels of diesel. Path C: price resolves the dispute. Positive tell: European diesel prices falling without any release. Exclusion tell: any coordinated release. CONFIRMS: An IEA- or G7-coordinated release that names diesel volumes by 15 November, with no US export restriction. REFUTES: A formal US restriction on diesel exports, or a European refusal with no release, by 15 November.
113-ie-2 · 2026-10-02 · O · Unresolved — Human corroboration arrives without override protection

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: The 95 percent corroboration threshold does not distinguish justified agreement from coerced signature. Cases requiring a justified override and protections for the decision maker must be observed.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: By 31 December 2027, audit or court records showing reviewers corroborating more than 95 percent of automated recommendations, with no statutory override protection. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 1 OCT 2026–31 DEC 2027 Human corroboration arrives without override protection OBS: California enacted SB 947, requiring human corroboration of AI-informed discipline and firing from 1 July 2027, and AB 1979, limiting clinical AI to an advisory role; the governor vetoed AB 2575, which would have protected workers who override AI recommendations. INF: The statute assigns the decision to a person without protecting the person who departs from the system’s recommendation. HYP: The disposition is toward corroboration working as a signature unless contracts or later statutes add protection for overrides. Release field: Path A: protection by contract. Positive tell: collective agreements or employer policies that protect overrides. Exclusion tell: upheld retaliation complaints from workers who overrode a system. Path B: review as signature. Positive tell: audit or litigation records showing reviewers corroborate more than 95 percent of automated recommendations. Exclusion tell: published override rates showing frequent departures. Path C: legislative return. Positive tell: an override-protection bill enacted in 2027. Exclusion tell: no such bill introduced in the 2027 session. CONFIRMS: By 31 December 2027, audit or court records showing reviewers corroborating more than 95 percent of automated recommendations, with no statutory override protection. REFUTES: Statutory override protection, or widespread contractual protection, in place by 31 December 2027, or published override rates showing frequent departures.
113-ie-3 · 2026-10-02 · O · Unresolved — The Horn’s rupture keeps an internal war at the edge of an interstate one

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Federal forces in or at the edge of Mekelle by 15 November with no independently confirmed Ethiopian–Eritrean engagement. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 2 OCT–15 NOV 2026 The Horn’s rupture keeps an internal war at the edge of an interstate one OBS: Ethiopia expelled ten Eritrean diplomats and closed its Asmara embassy on 1 October; Eritrea severed ties; Ethiopia and Egypt expelled each other’s diplomats; federal forces were reported about 40 kilometres from Mekelle; an Afar official alleged Eritrean troops in the region. INF: Breaking ties removes a channel between the two governments most directly involved, and each names the other as a cause of the fighting. HYP: The disposition is toward a continued federal advance with friction on the border below open interstate war through mid-November. Release field: Path A: interstate war. Positive tell: an independently confirmed engagement between Ethiopian and Eritrean units. Exclusion tell: an African Union envoy received in both capitals. Path B: settlement by force inside Ethiopia. Positive tell: federal forces entering Mekelle with no cross-border incident. Exclusion tell: renewed Tigrayan advances in Afar. Path C: mediated pause. Positive tell: an African Union– or IGAD-convened meeting of named delegations. Exclusion tell: further drone strikes on Addis Ababa or Mekelle. CONFIRMS: Federal forces in or at the edge of Mekelle by 15 November with no independently confirmed Ethiopian–Eritrean engagement. REFUTES: An independently confirmed engagement between Ethiopian and Eritrean forces, or a convened meeting of named delegations, before 15 November.
113-ie-4 · 2026-10-02 · O · Unresolved — Humanoid robots move from showroom to shop floor, or stay in the showroom

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: IDC data for the first half of 2027 showing the research and non-commercial share between 50 and 68 percent. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 2 OCT 2026–30 SEP 2027 Humanoid robots move from showroom to shop floor, or stay in the showroom OBS: IDC counted almost 25,000 humanoid robots shipped in the first half of 2026, up 432.1 percent; Chinese vendors supplied more than 95 percent; about 69 percent went to research, education, exhibition and government data centres, down from 83.8 percent; the average price fell to just under $30,000. INF: Volumes are growing fast from a small base, mostly into uses that do not yet pay for themselves in production. HYP: The disposition is toward a falling research share as prices drop, without a majority of units in paid work within the year. Release field: Path A: commercial turn. Positive tell: a research and non-commercial share below 50 percent in IDC’s first-half 2027 data, with named industrial customers placing repeat orders. Exclusion tell: a research share above 60 percent. Path B: showroom plateau. Positive tell: shipments growing less than twofold with the research share steady. Exclusion tell: growth above 200 percent. Path C: state-led absorption. Positive tell: government procurement in China taking the largest share of non-research units. Exclusion tell: private industrial buyers taking most of them. CONFIRMS: IDC data for the first half of 2027 showing the research and non-commercial share between 50 and 68 percent. REFUTES: A research share of 69 percent or more, or below 50 percent, in IDC’s first-half 2027 data.
114-ie-1 · 2026-10-05 · O · Unresolved — Brazil’s runoff starts from a first round the polls did not predict

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Flávio Bolsonaro winning on 25 October by fewer than five points of valid votes, with no formal challenge to the count. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 5–25 OCT 2026 Brazil’s runoff starts from a first round the polls did not predict OBS: Flávio Bolsonaro won 47.03 percent of valid votes and Lula 45.16 percent on 4 October. Datafolha’s last poll had Lula ahead by four points, and its last runoff scenario had him ahead 48 to 45. Augusto Cury took 2.89 percent and Renan Santos 2.25. INF: The first round consolidated most of the opposition vote early, and the instruments that simulate the runoff missed the order of the first round. HYP: The disposition is toward a runoff decided by a few points, with the first-round leader favoured and the published runoff simulations carrying less weight than usual. Release field: Path A: the order holds. Positive tell: post-first-round polls from two or more firms showing Flávio ahead. Exclusion tell: two or more firms showing Lula ahead. Path B: Lula recovers. Positive tell: a public endorsement of Lula by Cury, or two firms showing Lula level or ahead. Exclusion tell: Flávio above 50 percent of valid votes in polls from two or more firms. Path C: a contested result, which can accompany either. Positive tell: a formal challenge filed with the electoral court. Exclusion tell: a concession on election night. CONFIRMS: Flávio Bolsonaro winning on 25 October by fewer than five points of valid votes, with no formal challenge to the count. REFUTES: A Lula victory, a Flávio victory by five points or more, or a formal challenge to the count.
114-ie-2 · 2026-10-05 · O · Unresolved — The diesel release is delivered in barrels or in accounting

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: The later IEA statement supports the accounting concern, but the original terminal test requires cumulative new releases by February 2027. Neither a pledge nor an earlier pledge’s remainder is a delivered-barrel outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: By 2 February 2027, IEA or member accounts showing less than 100 million barrels of new releases, with part of the total attributed to earlier commitments. Bounded component checks are recorded as iea-1007. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O · RIPENESS: 5 OCT 2026–2 FEB 2027 The diesel release is delivered in barrels or in accounting OBS: The G7 agreed on 2 October to release 100 million barrels through the IEA over four months; the statement counts commitments already fulfilled and names no diesel volume or member shares. Friday’s reading looked for a coordinated release naming diesel volumes by 15 November, with no US export restriction. The diesel volume was not named, so Friday’s confirming condition is met in part, and the threat that reading kept in reserve was dropped. INF: The total and the route match France’s proposal. The diesel volume is unspecified. HYP: The disposition is toward a release that eases diesel prices for a few weeks, with a new physical volume smaller than the headline. Release field: Path A: delivered draw. Positive tell: IEA data showing diesel stock draws by member within the first 20 days. Exclusion tell: no product breakdown published by 30 November. Path B: netting. Positive tell: an IEA or member figure for how much of the 100 million was released before 2 October. Exclusion tell: published new volumes that sum to 100 million. Path C: renewed pressure. Positive tell: a new US request or export threat. Exclusion tell: the average US diesel price back below $5.61 a gallon by the end of December. CONFIRMS: By 2 February 2027, IEA or member accounts showing less than 100 million barrels of new releases, with part of the total attributed to earlier commitments. REFUTES: Published accounts showing 100 million barrels of new releases with a stated diesel volume.
114-ie-3 · 2026-10-05 · O · Unresolved — Tigray’s war continues from outside the captured capital

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: By 15 November, TPLF attacks outside Mekelle reported by two or more independent outlets, no signed cessation of hostilities, and no independently confirmed Ethiopian–Eritrean engagement. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 5 OCT–15 NOV 2026 Tigray’s war continues from outside the captured capital OBS: Government forces entered Mekelle on 3 October, and the TPLF said the same day that it had moved out and would fight on from other positions. On 4 October the leaders of Egypt, Eritrea, Somalia and Sudan’s army issued a joint declaration at New Alamein. Friday’s reading named federal forces entering Mekelle with no cross-border incident as one of three paths, and that tell has occurred. INF: The TPLF kept its forces by leaving, so holding the city does not end the fighting. HYP: The disposition is toward attacks from outside the city through mid-November, with the border with Eritrea tense and short of open war. Release field: Path A: war from the hills. Positive tell: TPLF attacks on roads or towns outside Mekelle. Exclusion tell: a signed cessation of hostilities. Path B: administered capture. Positive tell: an interim administration seated in Mekelle with electricity and flights restored. Exclusion tell: fighting inside the city. Path C: interstate war. Positive tell: an independently confirmed engagement between Ethiopian and Eritrean units. Exclusion tell: restored consular contact between Addis Ababa and Asmara. CONFIRMS: By 15 November, TPLF attacks outside Mekelle reported by two or more independent outlets, no signed cessation of hostilities, and no independently confirmed Ethiopian–Eritrean engagement. REFUTES: A signed cessation of hostilities, or an independently confirmed Ethiopian–Eritrean engagement, by 15 November.
114-ie-4 · 2026-10-05 · O · Unresolved — A portal’s final-submit step becomes law or yields to the statute

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: CONFIRMS treats no court order by the deadline as support for a legal-rule hypothesis. Procedural non-disposition does not establish the rule’s legal merits.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A Constitutional Court order before 4 November that leaves the final-submit requirement in force and the six lists excluded, or no order by that date. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: 5 OCT–4 NOV 2026 A portal’s final-submit step becomes law or yields to the statute OBS: The ANC failed to click “submit as final” in six municipalities before the 28 August deadline. The Electoral Court held that nominations stay provisional until the click. The commission says it built the step in 2016 and applied it to 2,274 candidates from 45 parties. The ANC complied in 251 municipalities. Local elections are on 4 November. INF: The Constitutional Court must decide whether a design choice in an electronic system can carry the force of a statutory condition. HYP: The disposition is toward the rule surviving, because it was applied across parties and the appellant met it almost everywhere, with the remedy for the six municipalities less certain. Release field: Path A: the rule stands. Positive tell: the appeal dismissed before ballots are final. Exclusion tell: an order to reprint ballots. Path B: the statute prevails. Positive tell: a holding that the click is not a legal requirement and an order admitting the lists. Exclusion tell: dismissal of the appeal. Path C: relief without a rule. Positive tell: the candidates admitted on fairness grounds with no holding on the click. Exclusion tell: a judgment that decides the click’s legal status either way. CONFIRMS: A Constitutional Court order before 4 November that leaves the final-submit requirement in force and the six lists excluded, or no order by that date. REFUTES: A holding that the click is not a legal requirement, or an order admitting the six municipalities’ lists.
115-ie-1 · 2026-10-06 · O · Unresolved — Iraq’s handover changes who directs the next security operation

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: By 31 March 2027, documented Iraqi-led operations with bilateral support and no restored coalition military mission in Iraq. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: NEAR · 6 OCT 2026–31 MAR 2027 Iraq’s handover changes who directs the next security operation OBS: The UK’s 30 September statement ends the coalition mission in Iraq and retains a bilateral defence relationship. INF: Departure changes the formal arrangement. Substantive transfer depends on whether Iraqi authorities can direct operations and obtain support under their own priorities. HYP: The disposition is toward Iraqi-led cooperation with outside support continuing through bilateral channels, rather than the disappearance of the outside relationship. Release field Iraqi direction. Positive tell: Published terms and an Iraqi-led operation identifying requested support. Exclusion tell: A foreign partner controlling the mission and its authorisation. Reconstituted dependence. Positive tell: Emergency return of a coalition formation or externally directed operations. Exclusion tell: Repeated operations under published Iraqi authority without emergency substitution. CONFIRMS: By 31 March 2027, documented Iraqi-led operations with bilateral support and no restored coalition military mission in Iraq. REFUTES: A restored coalition military mission in Iraq or independently documented external control of the operations described as Iraqi-led.
115-ie-2 · 2026-10-06 · O · Unresolved — Orbital compute earns value through a measured task rather than its location

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: By 6 October 2027, at least one independently documented outside workload with a specific advantage tied to processing near the data source. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: FAR · 6 OCT 2026–6 OCT 2027 Orbital compute earns value through a measured task rather than its location OBS: Axiom reports two orbital nodes and a ground cryptography migration. NASA’s manipulation platform remains a planned late-2027 mission. INF: The installed or planned hardware does not yet establish a service. The missing evidence is a task performed under real constraints and a user able to value it. HYP: The disposition is toward narrow useful services before broad competition with terrestrial infrastructure. Release field Specific service. Positive tell: A named outside user and published latency, recovery or avoided-transmission results. Exclusion tell: Only internal demonstrations with no service measure. Prototype persistence. Positive tell: Repeated milestone statements without an outside workload. Exclusion tell: A recurring paid workload with disclosed operating conditions. CONFIRMS: By 6 October 2027, at least one independently documented outside workload with a specific advantage tied to processing near the data source. REFUTES: Documented failure to sustain outside workloads under the stated conditions, or evidence that the proposed advantage disappears after complete communications and recovery costs are counted.
115-ie-3 · 2026-10-06 · O · Unresolved — Healthy-ageing implementation becomes observable when a policy gains a population record

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: The next comparable WHO progress report shows policy coverage increasing while the share able to track health over time remains below one third. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: FAR · 6 OCT 2026–31 DEC 2027 Healthy-ageing implementation becomes observable when a policy gains a population record OBS: WHO’s midpoint account records long-term-care policy coverage well above the share of countries able to track older people’s health over time. INF: Policy coverage cannot show whether a service reaches its intended population. A financed longitudinal record can reveal access, functioning and continuing unmet need. HYP: The disposition is toward policy expansion continuing faster than the monitoring needed to assess it, unless countries fund measurement with service development. Release field Coupled implementation. Positive tell: New funded longitudinal measures released alongside care investment. Exclusion tell: Strategy counts without access or functioning data. Formal expansion. Positive tell: Another increase in policy coverage with monitoring coverage little changed. Exclusion tell: A substantial increase in longitudinal monitoring with usable subgroup results. CONFIRMS: The next comparable WHO progress report shows policy coverage increasing while the share able to track health over time remains below one third. REFUTES: A comparable report shows at least half of countries able to track older people’s health over time with published service-access measures.
115-ie-4 · 2026-10-06 · O · Unresolved — A forest metric changes interpretation when its component transitions are retained

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: A comparative decision study finds that retaining component histories improves ecological targeting relative to the aggregate metric alone. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O · RIPENESS: NEAR · 6 OCT 2026–31 DEC 2027 A forest metric changes interpretation when its component transitions are retained OBS: The forest study finds lower global edge length produced through contrasting regional transitions. Its mapped period is 2000–2020. INF: A single direction in an aggregate can combine transitions with different effects. Decisions based only on the total risk rewarding the wrong local change. HYP: The disposition is toward local restoration and conservation decisions requiring edge change to be read with area and patch history. Release field Decision-relevant composition. Positive tell: A documented allocation or restoration choice changes when patch history is included. Exclusion tell: The same choice after the component history is withheld. Adequate aggregate. Positive tell: Held-out decisions and outcomes are comparable using only the total. Exclusion tell: Repeated local sign reversals that the aggregate misses. CONFIRMS: A comparative decision study finds that retaining component histories improves ecological targeting relative to the aggregate metric alone. REFUTES: A sufficiently powered comparison finds no decision or outcome benefit from retaining those histories across the relevant landscape types.
116-ie-1 · 2026-10-07 · O · Unresolved — Mechanistic attribution after a probe is reclassified

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: Independent discriminating comparisons support the nominated explanation in the stated contexts. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O Mechanistic attribution after a probe is reclassified The science item identifies a mechanism-specific claim within a defined experimental setting. The next comparison must ask whether the same explanation holds when relevant dependencies change. This orienting read directs that comparison without predicting clinical performance. Path A: independent comparisons favour the nominated explanation. Positive tells are distinct rival predictions, documented interventions and reproducible discrimination. A repeated endpoint without a distinguishing comparison is excluded. Path B: a contextual dependency favours another explanation. Positive tells are a specified boundary and a rival prediction that succeeds in that context. Unexplained variability alone is excluded. Ripeness: Six to eighteen months, April 2027–April 2028. A lack of immediate replication is unassessed. It is not a failure of the mechanism. Release field CONFIRMS: Independent discriminating comparisons support the nominated explanation in the stated contexts. REFUTES: The nominated explanation fails its distinct prediction in well-controlled comparisons while a rival account succeeds.
116-ie-2 · 2026-10-07 · O · Unresolved — A rating agency’s independence becomes assessable through contested decisions

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: A consequential rating and challenge remain inspectable, with a decision consistent with the stated process. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O A rating agency’s independence becomes assessable through contested decisions The AU listing supplies a scheduled launch and a stated ownership design. The inference is that usable independence additionally requires traceable decisions, publication and correction procedures. The read remains orienting because operational ratings and their use have not been established here. Path A: criteria constrain a disputed rating. Positive tells are a published rationale and a correction or appeal handled through named rules. A launch speech or favourable reception is excluded. Path B: an outside interest overrides those rules. Positive tells are a documented intervention and an unexplained departure from the method. Disagreement with a rating alone is excluded. Ripeness: Six to eighteen months, April 2027–April 2028, after a first inspectable rating or operational method. Earlier absence is unassessed. Release field CONFIRMS: A consequential rating and challenge remain inspectable, with a decision consistent with the stated process. REFUTES: A documented political or commercial intervention changes an assessment outside the stated procedure.
116-ie-3 · 2026-10-07 · O · Unresolved — Climate exposure can inform an activation rule without identifying avoided harm

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: An identified authority uses the specified warning to activate and document an available response. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O Climate exposure can inform an activation rule without identifying avoided harm UNICEF supplies a prospective exposure estimate. The inference is that its contribution to preparedness can be examined through a previously specified activation rule, while any claim about avoided harm needs an additional comparison. This separates the forecast’s use from its accuracy and from the response’s causal effect. Path A: resources are activated before local impacts under a published trigger. Positive tells are a dated rule, authorised release and delivery record. Retrospective claims of preparedness are excluded. Path B: the trigger is revised or blocked. Positive tells are an attributable revision or an unmet condition linked to funding or authority. A lower damage total alone is excluded. Ripeness: Three to six months, January–April 2027, for initial follow-up after the forecast season. Release field CONFIRMS: An identified authority uses the specified warning to activate and document an available response. REFUTES: The agreed actionable threshold is reached, but a documented funding or authority barrier prevents activation.
116-ie-4 · 2026-10-07 · O · Unresolved — Organisational complements can explain useful AI adoption

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: A credible matched or experimental comparison links the specified complement to an improvement in the AI-supported workflow. Bounded component checks are recorded as eca-ai. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O Organisational complements can explain useful AI adoption The ECA release identifies exposure, basic adoption and missing complements. The proposed causal relation is that changing tasks, management and data access can condition the benefit of using AI. This orienting read tests a causal hypothesis. The regional rates do not identify the relation. Path A: a specified workflow change improves performance with the application. Positive tells are comparable task records, an outcome measure and evidence about the changed complement. Tool-use growth alone is excluded. Path B: access expands without useful integration. Positive tells are persistent task bottlenecks and unchanged outcomes despite use. A short implementation delay is excluded. Ripeness: Twelve to twenty-four months, October 2027–October 2028, for a defined organisational comparison rather than a prediction about all firms. Release field CONFIRMS: A credible matched or experimental comparison links the specified complement to an improvement in the AI-supported workflow. REFUTES: An adequately observed comparison finds equivalent outcomes despite the proposed complement and application, or identifies another mechanism that explains the difference.
117-ie-1 · 2026-10-08 · O · Unresolved — A compressed rule is tested when the goal changes

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: A new controlled comparison favors the compressed account on prespecified goal-change predictions and reports its boundary. Bounded component checks are recorded as compression-preprint. They do not establish this complete criterion; see the source-check ledger for the exact scope.

READ: O A compressed rule is tested when the goal changes The reward-compression item motivates a transfer question. The mechanism proposed here is that a compact representation can support efficient choice while preserving only the distinctions used by the original goal. This is a live research comparison, not a prediction that one internal representation has already been identified. Path A: a compressed rule transfers selectively. Positive tell: preregistered responses follow its distinctive predictions after a goal change with stimuli held fixed. Exclusion tell: only speed or an unchanged-task endpoint is reported. Path B: an association-based account explains the changed-goal responses better. Positive tell: it succeeds on predictions the compressed account fails, with matched complexity and information. Exclusion tell: unexplained noise alone. The paths concern relative explanatory support and are disjoint under a prespecified model comparison. Ripeness: April–October 2027. Six to twelve months permits design and a new controlled comparison. No qualifying study in that interval leaves the question unassessed. Release field CONFIRMS: A new controlled comparison favors the compressed account on prespecified goal-change predictions and reports its boundary. REFUTES: A well-powered comparison favors the rival on those predictions. Equivalent fit refutes specificity, while leaving compression available as a descriptive approximation.
117-ie-2 · 2026-10-08 · O · Unresolved — A component temperature gain is tested in its service regime

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: A qualifying trial demonstrates the predicted component-level operating gain and identifies its conditions. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O A component temperature gain is tested in its service regime The cable item separates a measured physical input from a modelled service consequence. The proposed mechanism is that cooling can relax a thermal constraint when that constraint actually limits the operating component. Other electrical or network constraints can prevent that relaxation from becoming usable delivery. Path A: the thermal constraint is binding and usable load increases. Positive tell: a matched energized trial documents an additional safe operating load under stated weather and electrical conditions. Exclusion tell: an unenergized temperature comparison. Path B: another limit binds. Positive tell: equivalent thermal improvement with no usable load increase, alongside an identified electrical or system constraint. Exclusion tell: a trial too short or incomplete to assess load. These outcomes are disjoint for the same tested configuration. Ripeness: October 2027–October 2028. One to two years allows time for an independently inspectable operating comparison. Delay before this window is not deterioration. Release field CONFIRMS: A qualifying trial demonstrates the predicted component-level operating gain and identifies its conditions. REFUTES: A qualifying trial shows that the physical cooling gain does not create the predicted usable load in the tested configuration. This challenges transfer to that service regime, not the original temperature measurement.
117-ie-3 · 2026-10-08 · O · Unresolved — Institutional access is followed through the authority that can deliver a remedy

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: More timely adjudication is accompanied by independently inspectable implementation for the specified decisions. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O Institutional access is followed through the authority that can deliver a remedy The strategy release distinguishes adjudication from implementation. This chain proposes following a decision through the authority responsible for a remedy. Path A: the handoff is completed. Positive tell: decision-level records identify the duty, implementing actor and documented remedy. Exclusion tell: a throughput count alone. Path B: the handoff is blocked. Positive tell: an identified decision remains unimplemented after its applicable deadline, with a documented refusal or unmet authority condition. Exclusion tell: absence of public information alone. For a defined decision and deadline, implemented and documented blocked outcomes are disjoint. Incomplete records form an unassessed case. Ripeness: End of 2027 through the 2028 strategy cycle. The strategy horizon allows implementation to be assessed against actual decision deadlines. Release field CONFIRMS: More timely adjudication is accompanied by independently inspectable implementation for the specified decisions. REFUTES: Adjudication improves while documented implementation remains blocked for those decisions. That refutes a combined improvement claim, rather than demonstrating that judicial processing has no value.
118-ie-1 · 2026-10-09 · O · Unresolved — Sudan’s embargo is decided in December or rolled over a third time

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: The Council adopts a resolution that either extends the embargo beyond Darfur or renews the existing regime for a year. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O Sudan’s embargo is decided in December or rolled over a third time The Council has twice renewed the 1591 regime without changing it. The United States sought wider coverage, and six members preferred a year-long renewal limited to Darfur. The 11 December deadline forces a choice between these positions or a further neutral text. Path A: a substantive text is adopted. Positive tells are a circulated draft that states geographic scope and a renewal of twelve months. Public statements of concern without a draft are excluded. Path B: a third technical rollover. Positive tells are unchanged language and a period of three months or less. A brief extension of days to complete agreed drafting is excluded. Ripeness: Near. Nine weeks, to 11 December 2026. Release field CONFIRMS: The Council adopts a resolution that either extends the embargo beyond Darfur or renews the existing regime for a year. REFUTES: A third short renewal with identical language is adopted, or the regime lapses.
118-ie-2 · 2026-10-09 · O · Unresolved — South Africa rebuilds a compliant asylum intake or returns to court

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Reception offices accept the applications covered by the ruling under a published procedure, without the disorder of early October. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O South Africa rebuilds a compliant asylum intake or returns to court The directive is withdrawn and the Constitutional Court’s July order remains in force. The department has said it will start a new process. The order can be met through a staged procedure, or the gap persists until litigants seek enforcement. Path A: a replacement procedure is published. Positive tells are an appointment or quota system, a security plan for reception offices and a public explanation of what the ruling grants. A statement of intent to comply is excluded. Path B: enforcement litigation. Positive tells are a filed application citing non-compliance and a court-set deadline. Renewed street protest alone is excluded. Ripeness: One to four months, November 2026–February 2027. Release field CONFIRMS: Reception offices accept the applications covered by the ruling under a published procedure, without the disorder of early October. REFUTES: No procedure exists by the end of February 2027 and applicants covered by the ruling are still being turned away.
118-ie-3 · 2026-10-09 · H · Unresolved — Brazil’s runoff polls are compared with the count

Originating inference section · Test: dated event or document followup · Original READ: H

Stance basis: Retain the original H judgment: the author explicitly reduces the endpoint to one poll-versus-count comparison. It is not an R template claim. Other analysts could code the two branches O; report this boundary case rather than use it to infer severity.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: Both candidates’ valid-vote shares fall within two points of Datafolha’s final runoff estimate, converted to the same base. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: H Brazil’s runoff polls are compared with the count Datafolha underestimated the challenger in the first round and now shows him four points ahead inside a two-point margin. The configuration is a close race measured by an instrument with a recent directional error. The final step here reduces to a single comparison on one date, which is why the chain is tagged H: the count either falls within the poll’s range or it does not, and the read offers no field beyond that. Path A: the final Datafolha estimate and the count agree within two points on each share. A positive tell is a published note on any change in weighting. Agreement on the winner alone is excluded. Path B: the count departs from the estimate in the same direction as on 4 October. A positive tell is a larger challenger share than polled. A difference explained by blank and null ballots is excluded. Ripeness: Near. Sixteen days, to the count on 25 October 2026. Release field CONFIRMS: Both candidates’ valid-vote shares fall within two points of Datafolha’s final runoff estimate, converted to the same base. REFUTES: The challenger’s share exceeds the final estimate by more than the margin of error for a second time.
118-ie-4 · 2026-10-09 · O · Unresolved — A machine-led biological finding gains or fails to gain a functional result

Originating inference section · Test: dated event or document followup · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The terminal criterion remains unadjudicated at the cutoff: An independent group reproduces a defined biochemical activity and the released log shows unguided candidate selection. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O A machine-led biological finding gains or fails to gain a functional result A company reports that agents identified a candidate enzyme system in phage DNA. Function is undetermined and a priority claim is outstanding. Its status depends on bench evidence and on the search record. Path A: function is shown and the record is released. Positive tells are assay data in a preprint and published prompts. Favourable comment from senior scientists is excluded. Path B: the claim stays at the candidate stage. Positive tells are new candidate announcements without assays and silence on priority. A delay of under six months is excluded. Ripeness: Six to eighteen months, April 2027–April 2028. Release field CONFIRMS: An independent group reproduces a defined biochemical activity and the released log shows unguided candidate selection. REFUTES: Assays find no activity, or the log shows that staff specified the candidate families before the search.
119-ie-1 · 2026-10-11 · O · Unresolved — Fiscal pressure becomes a crisis only through a financing and propagation channel

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: the observed funding mechanism explains deterioration beyond policy-rate repricing. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O Fiscal pressure becomes a crisis only through a financing and propagation channel OBS: The markets section records higher US interest expenditure, small weekly Treasury-yield declines and earlier evidence of functioning markets. INF: fiscal exposure and current dysfunction are different states. HYP: a repricing becomes disruptive when funding constraints force asset sales faster than unlevered demand can absorb them. Release field: two discriminating paths. Orderly adjustment. Positive tell: repeated auctions clear with usable financing and stable repo terms even as effective fiscal costs rise. Exclusion tell: a yield decline alone, without funding evidence. Liquidity amplification. Positive tell: documented margin or repo tightening precedes forced sales and persistent depth deterioration. Exclusion tell: a fiscal forecast or correlated yield movements without those transmission observations. Ripeness: October 2026–January 2027 for funding channels; two to four quarters for fiscal transmission. CONFIRMS: the observed funding mechanism explains deterioration beyond policy-rate repricing. REFUTES: stable financing and absorption through a substantial repricing reject the near-term amplification thesis. Missing position data remain an identification limit. Preserve liquidity and match obligations while testing the mechanism; a crisis date is not implied.
119-ie-2 · 2026-10-11 · O · Unresolved — Cheaper information can strengthen judgment or displace the cost of verification

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: a controlled comparison shows that source-aware checking explains the joint gain. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O Cheaper information can strengthen judgment or displace the cost of verification OBS: The open-data tool expands ways to discover and inspect candidate evidence. INF: retrieval and substantive fit are separate tasks. HYP: the tool’s value depends on whether provenance and rejection remain inexpensive as the candidate set grows. Release field: two discriminating paths. Improved joint performance. Positive tell: faster defensible decisions with stable or lower consequential error under heterogeneous sources. Exclusion tell: time to the first answer alone. Transferred verification burden. Positive tell: faster retrieval accompanied by more downstream corrections or longer total completion time. Exclusion tell: additional checks that improve final quality without net delay. Ripeness: January–April 2027. CONFIRMS: a controlled comparison shows that source-aware checking explains the joint gain. REFUTES: equivalent gains and error rates occur without those checks on the same difficult cases. The appropriate next investment is a comparison of completed decisions, with rejected candidates retained.
119-ie-3 · 2026-10-11 · O · Unresolved — A service reform can support autonomy or transfer an unmet obligation

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: continuity and responsibility explain outcomes beyond discharge counts. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O A service reform can support autonomy or transfer an unmet obligation OBS: WHO names housing and other services as complements to community mental-health care. INF: changing the delivery location changes coordination requirements. HYP: continuity of support mediates whether the reform expands usable choice. Release field: two discriminating paths. Supported autonomy. Positive tell: sustained housing and care, greater participant control and no concealed increase in unmet family burden. Exclusion tell: fewer institutional beds alone. Displaced obligation. Positive tell: loss of care or housing after discharge, with tasks shifted to unsupported households. Exclusion tell: temporary disruption followed by documented recovery. Ripeness: October 2027–October 2028, assessed against actual local implementation. CONFIRMS: continuity and responsibility explain outcomes beyond discharge counts. REFUTES: comparable outcomes persist despite substantial measured differences in those complements. Follow the same people and missing cases; do not infer success from an administrative boundary crossing.
120-ie-1 · 2026-10-12 · O · Unresolved — A better proxy can improve decisions, redirect behavior or merely change the record

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: The field distinguishes measurement validity, behavioural redirection and record correction. REFUTES targets the redirection mechanism in a setting rather than refuting the entire three-path field; keep the mechanism-specific outcome separate from the overall research question.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: jointly observing scores, behavior and external outcomes separates these mechanisms. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O A better proxy can improve decisions, redirect behavior or merely change the record OBS: Robotic-skill assessment, clinical-audio grading and the corrected innovation ranking each depend on a defined reference. INF: changes in a measure and changes in its intended outcome must be observed separately. HYP: a measure becomes more useful when its relation to consequential outcomes survives use and adaptation. Release field: three discriminating paths. Useful measurement. Positive tell: improvement against an independent outcome persists in a new setting after users know the measure. Exclusion tell: better fit to the same training or consensus reference alone. Behavioral redirection. Positive tell: measured performance rises after incentives change while the independent outcome stalls, with observable effort shifting toward scored features. Exclusion tell: outcome noise without a behavioral change. Record correction. Positive tell: revised inputs change the score or rank while the underlying activity remains unchanged. Exclusion tell: a contemporaneous real improvement that independently explains the change. Ripeness: January–October 2027 for independent evaluations; January 2027 for ranking-correction propagation. CONFIRMS: jointly observing scores, behavior and external outcomes separates these mechanisms. REFUTES: stable external validity across changed incentives rejects the proposed redirection mechanism in that setting. Measurement error and selective participation remain limits.
120-ie-2 · 2026-10-12 · O · Unresolved — Common AI capability can expand a market or redistribute a fixed prize

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: concentration of actions mediates a differential return after accounting for demand. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O Common AI capability can expand a market or redistribute a fixed prize OBS: The World Bank reports rapid growth in Indian AI investment alongside continuing growth prospects. INF: the spending figure does not identify retained value. HYP: demand expansion and the similarity of competitors’ actions determine whether cheaper production supports broader gains or compresses private returns. Release field: two discriminating paths. Market expansion. Positive tell: adoption lowers quality-adjusted cost, new customers enter and output grows without equivalent margin erosion. Exclusion tell: revenue growth generated only by price inflation or acquisition of rivals. Competitive convergence. Positive tell: independently measured similarity in offers rises before entry congestion and margin compression in exposed markets. Exclusion tell: falling margins fully explained by weaker demand or higher input prices. Ripeness: two to four quarters after a specified adoption wave, through October 2027. CONFIRMS: concentration of actions mediates a differential return after accounting for demand. REFUTES: comparable returns persist across strongly different action correlations under matched demand and cost changes. Selection into adoption must be addressed before assigning causation.
120-ie-3 · 2026-10-12 · O · Unresolved — A shared institution becomes useful at the moment a handoff is required

Originating inference section · Test: comparative mechanism test · Original READ: O

Stance basis: Terminal prose retains alternative releases and an explicit possibility of refutation; coding uses the text, not a later outcome.

Timing: not yet due. The stated closing window extends beyond 12 October 2026. A qualifying event could resolve earlier, but absence before that endpoint is not refutation.

Outcome assessment: The confirming proposition requires a comparative test: specified changes in responsibility and access explain a reduction in boundary delays. No qualifying independent outcome check was completed for this chain; archive recurrence alone is not closure.

READ: O A shared institution becomes useful at the moment a handoff is required OBS: ASEAN power-grid work, Europe’s capital-market agreement and the cyber-crisis meeting programme identify coordination structures. INF: formal alignment and usable operating capacity are separate observations. HYP: clear responsibility, information access and resources determine whether the structure reduces delays across a boundary. Release field: two discriminating paths. Operational integration. Positive tell: documented cross-border transactions or exercises complete under the shared rules with fewer unresolved handoffs. Exclusion tell: a communiqué, meeting or nominal connection alone. Persistent fragmentation. Positive tell: comparable transactions repeatedly stall at the same legal or administrative boundary despite implementation. Exclusion tell: delay before the rules are due to operate, or disruption caused by an unrelated physical failure. Ripeness: April–October 2027, tied to each initiative’s actual implementation. CONFIRMS: specified changes in responsibility and access explain a reduction in boundary delays. REFUTES: equivalent performance without those changes rejects their claimed necessity. These are analogous tests across different institutions, not one pooled causal estimate.

Source checks and reproducibility

Primary sources were retrieved on 12 October 2026. They support only the stated components or research anchors. Ten checks concern outcome follow-up, including two Held event criteria; five ground proposed research. Publisher access to the 2026 compression article failed, so the dated 2025 original preprint is explicitly used as historical evidence.

SourceDateRole and limit
Federal Reserve: September FOMC statement2026-09-16A unanimous quarter-point increase to 3.75–4.00 percent occurred. This establishes the decision, not whether data moved market odds more than speeches.
ECB: September monetary policy decision2026-09-10The ECB raised its three key rates by 25 basis points and cited continuing inflation pressure. The decision does not identify the relative causal influence of energy, core inflation, and institutional communication on market pricing.
US Treasury: G20 Chair’s Statement2026-09-01A dated chair’s statement exists, with a footnote identifying China’s dissent on four paragraphs. Its requested actions and reporting work do not fit an inference from silence. Implementation and subsequent migration of negotiating authority remain separate tests.
IEA: member-government stock-release statement2026-10-07The IEA links approximately 100 million barrels to stocks pledged but not yet released under the March action. This supports an accounting distinction between previous commitments and delivered barrels; it does not establish cumulative physical delivery by February 2027 or a final diesel allocation.
IEA: G7 energy-security meeting2026-10-02The IEA reports the meeting and continuing diesel pressure. A policy response was discussed; realised supply and price effects require later observations.
NASA: Roman instrument activation and coronagraph checks2026-09-15Instrument activation is a later commissioning milestone. It does not by itself establish calibrated survey science, the economic value of a robot deployment, or a cross-program relation between reporting discipline and outcomes.
China NBS: September purchasing managers’ indices2026-10-08The official follow-up provides separate production, orders, employment and size measures. The Chinese branch can be inspected; a pooled mechanism joining Chinese manufacturing and South African investment is not established by that one release.
ARPA-H: autonomous stroke-intervention programme2026-08-06The programme specifies staged demonstrations and a separate simulation and validation testbed. It motivates a distinction between execution capability and verification; it is not evidence of deployed clinical autonomy or durable human complementarity.
Ann, Liu and Tan: The Interaction Tax2026-08-24The authors report 11 verifier-scored optimisation tasks under matched budgets. Full-solution communication reduced solution diversity; critique benefits depended on whether violations were detectable. Generalisation to venture competition and organisational work remains a hypothesis.
World Bank: Making AI Work in Europe and Central Asia2026-10-06The release distinguishes exposure from adoption and names managerial skills, interoperable data and computing capacity as constraints. Its regional description does not causally identify any particular organisational complement.
Molinaro and Collins: reward-function compression, original preprint2025-09-08The original preprint proposes and tests a compression account across six behavioural experiments. It is historical primary evidence for the idea; it is not a new 2026 observation or direct evidence of LLM internal representations.
Accenture and Anthropic: embedded evaluation partnership2026-09-18The companies announced an embedded evaluation arrangement. The announcement verifies the organisational form, not evaluator independence in contested decisions or an exercised right to delay release.
Environment Agency: drought report, 25 September–1 October2026-10-05The report records 767 abstraction restrictions, up from 533, while reservoir storage continued to fall. This meets 102-ie-1’s event criterion of an institutional restriction change before the end of October. It does not establish the causal model across the other cases.
Environment Agency: RPS 382, abstraction outside the authorised period2026-10-01Published on 1 October and clarified on 7 October, the statement describes conditional enforcement discretion outside authorised abstraction periods. It is an additional observed institutional response, not a repeal of licensing requirements or a demonstrated recovery of storage.
African Union: defence committee report and dated Council handoff2026-09-21The AU reports adoption of the committee report for submission to the Executive Council on 27–28 January 2027. This is a published institutional handoff with a named recipient and date. It does not establish a new budget, executed security operation or costly transfer.