Hiyma
LiveDesksPricing
Terminal desks
8 live · 5 building · opens in the terminal
GWRWar RoomliveConflict, shipping, chokepoints and risk zones, read spatially.GODGod-ViewliveA 3D picture of movement across the planet, including ships that go dark.NEWSNews IntelligenceliveThe global firehose as events, entities, places and market relevance.FINMarkets CommandliveEquities, commodities, FX, rates, crypto and macro on one clock.EQEquities CommandliveFrom ticker to tape: filings, events and market structure.OPTOptions CommandliveFlow, volatility, unusual activity and positioning behind the move.MOBMobility & LogisticsliveAir, sea, space and land: what is moving, what stopped, what is late.WSWorkspaceliveOne canvas: arrange any instrument around a single situation.CRYPTOCrypto CommandbuildingThe majors on a 24/7 clock: price, leverage and flows.ARMSArms IntelligencebuildingArms sales, transfers and contracts as they are announced.HEDGEBallastbuildingYour business in one terminal: exposure, drivers, hedges, stress tests.SEMISemiconductor TerminalbuildingThe AI industrial stack: labs, silicon, power and unit economics.REOReorderbuildingEight bets that the world’s plumbing is being rewired.
Terminal desks

Each desk answers one kind of question and reads the same live picture as every other, so a headline, a ship and a price move can be followed across them without changing tools.

Request terminal access
Try it, no keyAsk the live picture a questionEvery figure has a receiptSee what one carriesCompare plansFree seat to sovereign
By the question you are paid to answer
Investors & macro desks“Does the story agree with the price?”Geopolitical & country risk“Where is pressure building, and what does it touch?”Supply chain & trade“Which routes and counterparties are exposed?”Sanctions & compliance“Who is designated, and who is moving near them?”Analysts & investigators“Can I show my working?”AI agents & automation“Can my agent trust what it reads?”
By sector
Government and DefenseFinancial MarketsEnergy and UtilitiesSupply Chain and LogisticsIndustrial and ManufacturingTechnology and CyberInsurance and Risk TransferHumanitarian and Public InterestCorporate Security and Executive RiskResearch and Advisory
By workflow
Strategic WatchOne live watch lane when a country, route or book turns.Exposure MappingTrace an event to the assets, suppliers and positions it touches.Scenario LabStress a what-if before it happens.Operational BriefingA cited brief from the same picture the desks show.Compliance and DiligenceDesignations beside the entities and ships they touch.Command OperationsA shared operating picture for a crisis cell.
100 buyer profiles Find your deployment fit
See every use case
Version 2
Verifiable intelligence

The doctrine, how it is enforced, and the evidence. Every figure in the paper is measured and sealed, and any one of them can be checked from inside the paper.

Read the paper Check a figure
Part I · The argument
0Abstract1The hollow-data problem2Doctrine: never fabricate13What is not yet true
Part II · How trust is enforced
3Where every read comes from4The data spine5Enforcement, written down6Knowing what we know7The trust layer8AI under the same rules9The trust boundary
Part III · Evidence & road ahead
10Every noun is a URL11Data strategy12Evidence14Roadmap15Check the paper yourself16Glossary & references
RelatedSecurityComplianceDocumentationSystem status
Uplink
UNCLASSIFIED // PUBLIC DRAFT
DOC_REF: HIYMA/WHITEPAPER/v1.0

Whitepaper

Verifiable intelligence on an honest warehouse: the doctrine, the machinery that enforces it, and the evidence — every figure measured from the code and sealed.

Contents
  1. 0Abstract
  2. 1The hollow warehouse
  3. 2Doctrine: never fabricate
  4. 3The request path
  5. 4The warehouse spine
  6. 5Enforcement as code
  7. 6The epistemic layer
  8. 7The Trust Layer
  9. 8AI under the same doctrine
  10. 9The trust boundary
  11. 10Every noun is a URL
  12. 11Data strategy
  13. 12Evidence
  14. 13What is not yet true
  15. 14Roadmap
  16. 15Verify this paper
  17. 16Glossary & references
Contents
  1. 0Abstract
  2. 1The hollow warehouse
  3. 2Doctrine: never fabricate
  4. 3The request path
  5. 4The warehouse spine
  6. 5Enforcement as code
  7. 6The epistemic layer
  8. 7The Trust Layer
  9. 8AI under the same doctrine
  10. 9The trust boundary
  11. 10Every noun is a URL
  12. 11Data strategy
  13. 12Evidence
  14. 13What is not yet true
  15. 14Roadmap
  16. 15Verify this paper
  17. 16Glossary & references
Version 2.0
COMMIT259feb2
MEASURED2026-11-21
SEALED FIGURES38
ROOTddb9f461c574ebe7…

Precompute trust, stream deltas, never fabricate.

OpsHiyma is an operational-intelligence platform that fuses geopolitical events, markets, sanctions, mobility, humanitarian, energy and space data into one working picture. This paper describes the one constraint the platform is built around — it does not serve a number it cannot account for — and shows how that constraint is enforced by code rather than promised by prose. Every figure below was measured from the repository by a script anyone can run, and the set of figures is sealed under a Merkle root you can check.

§ 0

Abstract

Intelligence platforms fail in a specific, quiet way: the warehouse is emptier than the dashboard, and the gap is filled with plausible numbers. OpsHiyma’s own audits in mid-2026 found exactly that failure in its own code. The response was not a cleanup but a doctrine with teeth. Every served value carries a provenance envelope that states its source, its as-of time, and one of six honesty states. A user request never waits on a vendor; desks read canonical warehouse atoms or dated serving envelopes. A fabrication linter and a request-path linter run in CI with ratcheting debt ledgers that can only go down. On top of that floor sits an epistemic layer — content-hashed evidence, an Analysis of Competing Hypotheses kernel that refuses to rank an under-filled matrix, prediction ledgers scored by Brier against baselines, and a Trust Layer that reconciles disagreeing sources and issues Merkle certificates.

The paper closes with the evidence: 509 warehouse models, 1458 API routes, 289 connector manifests (224 keyless), 19853 unit tests, and the live row counts of the warehouse at the moment you loaded this page — followed by a ledger of what is not yet true, because a paper about honesty that hid its own gaps would refute itself.

§ 1

The hollow warehouse

A data platform’s atom is a scalar cell, and a cell has no memory of where it came from, how sure anyone is, or who disagrees. Fill the store slowly and render it fast, and the gap gets filled with invention.

The failure is easy to describe and hard to see. A route needs a number; the table behind it is empty or stale; the developer reaches for a random walk, a hardcoded tape, a “representative” scenario. The panel renders under a live-looking header, the demo lands, and the invented value becomes indistinguishable from a measured one. Nothing crashes. That is the problem: fabrication is a silent success.

OpsHiyma lived this. A launch-readiness audit on 9 June 2026 counted roughly twenty endpoints and engines serving random or hardcoded data styled as real, a provenance module that was complete and used by one engine, and no scheduled ingestion in production. A systems audit three weeks later put the figure at about a third of the then-130 API routes serving some synthetic data, found fourteen stray database clients that would exhaust Postgres connections at the second Cloud Run instance, and found the best-engineered code in the repository — the Trust Layer — imported by nothing. Those audits are in the repository, unedited, under docs/launch-readiness-audit.md and docs/performance-and-ip-master-plan.md.

Five amnesias

Generalized, the local bug is the industry’s architecture. Four generations of warehouses changed where storage and compute live and never changed the atom. A modern warehouse has five kinds of amnesia:

  1. Provenance amnesia. Lineage answers which job produced a table, not what evidence and confidence sit behind one value.
  2. World-model amnesia. “Apple” in sales, “AAPL” in market data and “Apple Inc.” in the news table are three strings, not one entity.
  3. Single-source amnesia. Conflicts are resolved at write time into a golden record, and the disagreement — often the signal — is discarded.
  4. Scalar amnesia. Most real quantities are distributions with uncertainty; a cell flattens them into false precision.
  5. Passivity amnesia. The store waits to be queried. It does not decay stale beliefs, detect new contradictions, or re-check itself.

AI makes each of these worse, because model-generated values now enter stores looking exactly like measured ones, and because agents are about to act on those stores with no way to tell a corroborated measurement from a single-source guess. The rest of this paper is the answer OpsHiyma was forced to build for the hardest version of the problem — adversarial, contradictory, multi-source geopolitical data — and the reason we think that answer generalizes.

§ 2

Doctrine: never fabricate

Random numbers dressed up as live intelligence are never acceptable.

That sentence is the header comment of lib/provenance.ts, and everything else follows from it. A route either returns real data with a status of live, cached or stale, an explicit no-data envelope when the warehouse or source is empty, an error when the source failed, or simulated when a value is modeled rather than measured — in which case it is also flagged modeled: true and must be labelled as such wherever it is drawn.

live
fetched or computed from a real source this request
cached
real data served from a server cache
stale
real data older than its freshness target
simulated
modeled, not measured — always flagged modeled:true
no-data
the source or warehouse returned nothing — honest empty
error
the source errored; the value is unavailable
Every served value is exactly one of these. Three are real data at different ages, one is a labelled model, two are honest absence. There is no seventh state for a plausible number.

The envelope

FieldMeaningWhy it exists
sourceHuman-readable source name, with an optional click-through URLA number with no named source is a rumor
asOfWhen the underlying data was produced or observed; null when unknownTime honesty: the world’s clock, not ours
fetchedAtWhen this server fetched or computed the valueCapture time must never masquerade as observation time
statusOne of the six states aboveThe reader can tell real from modeled from absent at a glance
confidence0..1 where a calibrated number existsOptional on purpose — an uncalibrated confidence is decoration
modeled / noteTrue for simulated values; a qualifier for tooltipsSimulated and no-data must be visually unmistakable from live

Two of those fields carry most of the doctrine. asOf versus fetchedAt is the difference between “this price is from 16:00 yesterday” and “we looked at 09:12 today”; an options change once let capture time pollute the as-of stamp of a desk and was treated as a correctness bug, not a cosmetic one. And no-data is a first-class answer: a panel that renders an honest empty is doing its job, and the UI has a dedicated vocabulary for it — a provenance badge, a simulated tag, and a full-width prototype banner for surfaces that are illustrative by design.

Adoption, measured

971 of 1458 route handlers (67%) import the provenance module today. The paper does not round that up. The remainder is covered by the gates in §5, which stop the number of unlabeled fabrications from growing while the envelope is adopted route by route.

§ 3

The request path

Vendor I/O belongs in ingest. The user GET reads the warehouse. A provenance-honest live fetch is still a violation.

The doctrine froze on 3 September 2026 (docs/warehouse-request-path.md) after the options desk proved the shape: read Postgres, return immediately, and refresh the vendor — if at all — after the response has been sent. Four rules, applied to every desk:

  1. The handler never waits on a vendor.
  2. History is warehouse-only. A past date is served from stored rows or not at all.
  3. Today may refresh in after(), a connector, or a worker — never in the GET.
  4. Missing data is an honest empty, not an upstream round trip.
Upstream sourcesgovernment open dataexchanges, wires, sensorsmanifest: auth · freshnessConnectorssince-cursor watermarksparse → RawRecordpayloadHash (sha256)Canonical atomsEvent · Series · ObservationGeoFeature · Entity · RelationshipClaim · Evidence(contentHash)IngestRun audit row per runREQUEST-PATH BOUNDARYno vendor I/O crosses to the rightServing envelopesDeskEnvelope by UTC dateGET returns the rowafter(): reassemble todayDesksobjects = URLscopilot toolsAPI + badgesProvenance envelopesource · asOf · fetchedAtstatus · modeled · noteon every served figureGates in CIfabrication ratchet · request-path ratchetseries-key contract · nav + login drift guardstypecheck (strict) · unit tests · build
Vendor I/O happens only left of the dashed boundary. Everything a user GET touches is a warehouse read or a dated serving envelope; today's envelope may be reassembled after the response is sent, never before.

Serving envelopes

Heavy desks do not recompute on every request. They store a dated JSON product in a DeskEnvelope row keyed by desk, UTC session date and schema version, and the GET returns that row. Only today’s row can be stale; when it is, reassembly from canonical atoms is scheduled after the response and deduplicated per desk so a sixty-second poll from many tabs cannot stack rebuilds. A schema mismatch always refreshes. 20 desk schemas are registered this way today, each with its own freshness budget:

From docs/warehouse-request-path.md §Phase 3 and lib/desk-envelope/keys.ts
DeskEnvelope keyFreshness budget
Truth Boardfusion:truth-board:{lookback}60 s
Escalation board / regionescalation:board · escalation:region:{key}15 min
Commodity War Roomcomwar:board60 s
Country dossiercountry:{iso2}6 h
MaritimeLogisticsSnapshotworker ingest; the GET is read-only
GET /api/desk?date=YYYY-MM-DDdefault: UTC todayread DeskEnvelope(deskKey, sessionDate, schema)index: deskKey, assembledAt descrespond nowpayload + provenanceorigin: warehouse | warmingstale + today?schema mismatch → refreshhistory → warehouse-onlyafter(): assemblefrom canonical atoms onlyone in-flight rebuild per keyupsert envelope(deskKey, date, schemaVersion)coverage recorded beside itA vendor is never awaited on this path.Missing data is an honest empty.
The handler reads the dated row and returns it. Only today's row can be stale; when it is, reassembly from warehouse atoms is scheduled after the response, deduplicated per desk so a poll stampede cannot stack rebuilds.
The request-path ledger, measured

The gate that enforces these rules keeps a ledger of tolerated legacy offenses. At commit 259feb2 it lists 0 files and 0 offenses — every request-path vendor fetch has been paid down, and a new one fails the build.

§ 4

The warehouse spine

Everything a desk shows is a read of a canonical atom that a connector wrote, behind a raw record that was hashed on arrival.

A connector is a directory with a manifest — its class, freshness target, pull type, auth, endpoints, sync strategy and target tables — and a parser. The ingestion engine runs it, writes an IngestRun audit row with status, cursors in and out, fetched and upserted counts, and observed lag, drop rate and jitter, then persists every original payload as a RawRecord with a SHA-256 payload hash. Raw records carry two independent unique keys, external id and payload hash, and the engine resolves against both before writing — a re-run of any snapshot connector used to violate the second key and take the connector down after its first successful run.

Canonical atoms

AtomKey disciplineWhat it remembers
Eventsource + externalId; optional clusterIdoccurredAt and t0Utc, latency from source to ingest, first-seen source, truth / impact / actionability scores
Series → ObservationSeries unique on (provider, dataset, seriesKey); Observation unique on (seriesId, ts)ts is when the world was in that state; reportedAt is when the source published or revised it — replayable as reported
GeoFeatureunique on (source, layerKey, externalId, validFrom)GeoJSON geometry with validFrom / validTo, a shaded metric and unit
Entity · Relationshipnormalized name, aliases, identifiers, match confidenceThe world model the events, sanctions, claims and exposures hang from
Claimsubject – predicate – object with qualifiersstatus ACTIVE / DISPUTED / RETRACTED / CONFIRMED, and the contradicting claim’s id
EvidencecontentHash unique (SHA-256 of the original)The citation spine: two citations of one wire story are one document

Labelling happens once, at ingest. The headline labeller replaces substring matching — where “war” matched “software” and “gas” matched “Gaza” — with a word-boundary country tagger, stem-and-phrase topic matching, an auditable tone lexicon, and an urgency that is source-declared only: a wire marker at the head of the title raises it, and no keyword ever does. The same module compiles into the standalone ingestor and the web app, so there is exactly one labeller.

The series-key contract

Three times the platform shipped a reader pointed at a warehouse key that did not exist, and each time the surface went quietly blank rather than loudly wrong — once the Truth Board’s reality leg matched zero of 1,309 series, so its headline divergence score had never been computed in production. No type check catches a string constant that matches nothing. The warehouse key contract gate now makes every declared reader key resolve to a real row, or the build fails.

Warehouse models
509
repo@259feb2· 2026-11-21
Declared indexes
1,041
repo@259feb2· 2026-11-21
Migrations
301
repo@259feb2· 2026-11-21
Models with orgId
216
repo@259feb2· 2026-11-21
§ 5

Enforcement as code

A promise you cannot lint is a hope.

The doctrine is only interesting because it is enforced where code lands. Every push runs 0 gates in continuous integration, read from the workflow file at measurement time:

Two ratchets

The fabrication gate has two detectors. The first flags Math.random() in a data-serving path unless the draw declares its intent — a file-level annotation for synthetic generators whose output is labelled simulated at the route boundary, or a line-level one for ids, nonces, jitter and layout. The second flags an array of three or more object literals carrying intelligence-shaped keys — price, volume, spread, probability, impact, score — inside a module that reads no data source: the literal tape of invented prints under a live-looking header, which no random-number detector can find.

When the gate was widened to the component tree it found real debt that could not be paid in one change, so it ratchets: each file may carry at most the offense count recorded in a ledger; more fails the build; fewer also fails, asking you to ratchet the ledger down; a file not in the ledger starts clean. Debt never silently reappears once paid. The request-path gate has the same shape for outbound vendor I/O and value imports of vendor SDKs in request-path modules.

Files in the fabrication ledger
0
repo@259feb2· 2026-11-21
Tolerated fabrication offenses
0
can only decrease
repo@259feb2· 2026-11-21
Tolerated request-path offenses
0
repo@259feb2· 2026-11-21
Served-path files with Math.random()
20
incl. annotated ids / jitter
repo@259feb2· 2026-11-21

Drift guards

Three tests walk the app/ directory and compare it with a registry, so coverage is a property of the data rather than of whoever last edited a navigation component. The nav-tree test fails when a page exists that no navigation surface lists, or when a nav row points at no page — before it, twenty-seven shipped pages were reachable only from inside their parent and two had no inbound link anywhere. The login-gate test holds the open tier and the public route group in lockstep, so no page becomes readable by strangers — or stops being readable — by accident. A third fails when a page under that group is not registered in the shell map, which this paper triggered on its first render: it drew the desk chrome over its own text and could not scroll. And the paper’s own test re-measures the repository and fails when the committed snapshot drifts more than ten percent from the tree — the paper may age, but it may not lie.

Underneath, TypeScript runs with 6 strictness flags (strict, exactOptionalPropertyTypes, noUncheckedIndexedAccess, noImplicitOverride, noFallthroughCasesInSwitch, verbatimModuleSyntax) and 19853 unit tests across 2371 files. The suite compiles pure modules directly rather than mocking them, including 40 connector parsers, so a vendor that changes its payload shape is caught against a captured fixture before the change can reach the warehouse.

§ 6

The epistemic layer

The honesty floor says what a number is. The epistemic layer says what we believe, on what evidence, with what confidence — and refuses when it cannot.

The long-run thesis (docs/epistemic-warehouse-thesis.md) is to change the atom: from a scalar cell to a Verifiable Claim — subject, predicate, object; a value or a distribution; an epistemic class from measured to synthetic; a content-addressed evidence chain; corroboration and contradiction; calibrated confidence with a trajectory; and a proof. What exists today is the working seed of that atom, built where the intelligence use case forced it.

Evidence that refuses

Evidence had held zero rows since it was designed until the Assessment desk gave it a writer. That writer enforces four refusals rather than four conventions: it never mints evidence from the platform’s own computed output, because citing a score we calculated as evidence for a judgment we are calculating is how a provenance platform starts fabricating; it never cites the future, because a citation carries a knowledge cursor and a row ingested after it is refused as not-yet-known — the writer records why: on this warehouse the median displacement record lands 122 days after the event it describes, so a replay without that cursor has look-ahead bias in every run; the only source it may create is an analyst upload; and it deduplicates by content hash, not by row.

Analysis of Competing Hypotheses, done the right way round

The Assessment desk runs Heuer’s method as a pure kernel — no I/O, no dates, no randomness, deterministic for a given matrix — with four load-bearing rules. Consistency never enters the ordering; only weighted inconsistency does. Diagnosticity multiplies into the ordering key, so evidence that says the same thing about every hypothesis cannot move the ranking — that invariance is the acceptance test. “Not applicable” is not zero; it changes the coverage denominator. And an under-filled matrix refuses to rank. Every constant lives in one parameter block whose hash is written onto the seal, so a published ranking can be replayed years later and proven.

Matrixhypotheses × evidencecells: CC C N I II NANA ≠ 0 (changes coverage)Row weightAdmiralty reliability A–Fcredibility 1–6, relevancedeception flags lower weightOrdering keyΣ weighted inconsistency× diagnosticityconsistency: excludedRank or refusecoverage < floor → REFUSEties stay tiesparams hash in the sealICD 203 confidence (low / moderate / high) is written on four axes — sourcing, logic, assumptions, gaps — and carries no probability.Likelihood uses the seven-band yardstick. The two are never conflated; a test pins it.
Consistency never enters the ordering; only weighted inconsistency does, multiplied by diagnosticity, so evidence that says the same thing about every hypothesis cannot move the ranking. An under-filled matrix refuses to rank rather than guess.

Around the kernel sit six models — Assessment, Hypothesis, EvidenceLink, AssumptionCheck, Judgment, ResolutionCriterion — with append-only revisions, a status machine from draft to published, superseded or retracted, and a blocked state whose payload names the missing collection: which lane, which connector, what its last run said, and what would unblock it. A model may draft an assessment; a human commits it, and model prose is never rendered as product — only the structured fields it filled.

Ledgers that are graded by the world

A political-risk desk’s claim that it can say something about a vote before it happens is untestable unless it is written down before the vote in a form that cannot be argued with afterwards. The election call ledger records, one day before polling, the crowd’s normalized probability vector, the poll average over the same outcomes when one exists, and a uniform prior over the outcome count — the floor a model must beat to be worth anything. Scoring is multi-class Brier, with a skill score against each baseline, and nothing is graded until the result is known from the world — a declared winner or a settled contract — never from the market’s own final price, because grading a market against itself always flatters it.

Three legs, one board

The Truth Board reads what sources say (narrative tone from GDELT), what crowds price (Kalshi and Polymarket contracts), and what measurement shows (warehouse series), and treats the spread as the signal. Its first working version compared raw levels and immediately produced five of six topics “divergent” at the same values every time, because news vocabulary is structurally negative and most geopolitical contracts are tails: differencing a 60 and a 12 yields a constant, not an observation. Each leg is now standardized against its own distribution across the topics on the board, so a constant offset cancels and the question becomes the one the desk actually asks — relative to how alarmed the story is about everything else, is the money unusually calm about this? A row is divergent only when some leg is a genuine outlier, not merely the largest member of a quiet set.

The same discipline runs through the regional escalation nowcast — nine weighted inputs, a confidence that is the coverage of the inputs applicable to that region, and a NO_DATA state a Gulf reading is not penalized into for lacking a European grid — and through the political-risk tools, which return “watched, not scored, because …” for an unmeasured country rather than a plausible number.

§ 7

The Trust Layer

A standalone, warehouse-agnostic library that audits a dataset for honesty, scores it, reconciles disagreeing sources into one belief, fingerprints synthetic data, and issues a verifiable certificate.

lib/trust-layer/ has no dependency on the platform’s schema or runtime; it consumes a dataset profile from an adapter and could be lifted into its own package unchanged. Its pipeline runs audit → trust score → certificate, and reconciliation beside them:

StageWhat it doesDetail that matters
AuditPer-column honesty findingsmissing provenance · stale-as-fresh · fabrication fingerprints · single-source critical fields · low epistemic class
Trust score0–100 with a letter gradeAggregates findings, epistemic quality and contamination with hard caps; the demo dataset scores 69 / D
ReconciliationN source claims → one belief stateTrust-tier weights S 1.0 · A 0.8 · B 0.6 · C 0.4 · D 0.2; epistemic weights measured 1.0 → synthetic 0.1; MAD outliers; consensus unanimous / majority / split / dissent / single / none
Contamination0..1 synthetic-data scoreConstant or low-distinct values, excessive float precision, histogram flatness, Benford first-digit deviation, templated text
CertificateSHA-256 Merkle proofLeaf and node hashes are domain-separated (RFC 6962 style) so a leaf can never be replayed as an internal node

The domain-separation detail is there because an audit found it missing: without the prefixes, an internal node could be presented as a leaf with a valid inclusion proof — the classic second-preimage forgery. It was fixed before the certificate was used for anything that mattered, and the seal on this page uses the fixed code.

Where it is wired, measured

162 modules outside the library import it: the admin trust page, its audit API, and the Truth Board’s leg-trust scoring. 4 dedicated test files cover reconciliation, contamination, certificates and the audit. Adapters exist for in-memory profiles; the Postgres adapter that would point it at the platform’s own warehouse is the next build, and until then this paper does not say the Trust Layer certifies the serving path.

§ 8

AI under the same doctrine

A model is an untrusted source like any other: it is routed, cross-examined, hashed, gated, and given tools that refuse to guess.

The Sovereign Intelligence Fabric routes each task to a model through one task router with adapters for Anthropic, OpenAI, xAI and local Ollama models, so callers speak a neutral message and tool vocabulary and the provider is a routing decision. The Adversarial Consensus Engine sends the same prompt to several models, builds an agreement matrix, detects divergences — which it treats as the interesting output — computes each model’s bias signature, scores confidence by track record and agreement, and produces a fused answer with provenance. The provenance chain hashes each model’s reasoning with SHA-256 into a record of which models were consulted, what they agreed on, and what the alignment rules said. Those alignment checks and the human approval queue write to one shared, queryable gate log — the existing audit trail — rather than each inventing its own.

The tool loop

The copilot at /api/copilot is the first endpoint that hands a model a tools array. The loop calls the model with the available tools, executes what it returns, appends the results, and calls again; it terminates when the model stops calling tools, at a hard iteration ceiling, or by returning an executor error to the model instead of aborting. Only tools that require no approval — the read and analysis set — are exposed, and the gate is enforced twice: withheld from the schema list, and refused by the executor even if a model names one. The transcript of every call and result is returned with the answer, inside a provenance envelope.

Desks are tool-first. The model does not answer “how unstable is Romania” from memory; it calls the country reading, which returns every leg with its reason and the coverage-derived confidence, or the reason the country was held back. An answer can therefore cite its evidence, and an unmeasured country produces a refusal rather than a plausible number. The Assessment desk applies the same rule to the model’s own prose: a model may draft, its transcript is kept beside the validator’s rejection log, and none of that text is rendered as product.

Admitted limits

The consensus engine measures agreement with token-overlap (Jaccard) similarity, and the code says so in a comment. Embeddings are the planned replacement and the single highest-leverage upgrade to the engine. This paper treats a comment that admits a shortcut as better than a diagram that hides one.

§ 9

The trust boundary

Default-deny at the edge, a signed identity that cannot be forged, a tenant guard that lives in one place, and a metering seam that never slows a request.

  • Default-deny proxy. Every API route requires an authenticated identity; elevated paths — admin, agents, ingestion, premium and analyst surfaces, geocoding, the mutating control-plane routes — add a role requirement. The genuinely public API is two probes: health and readiness.
  • Every desk needs a session; the open tier is a declared list. One module names what renders without signing in — the login page, static assets, and the marketing and legal pages under the public route group, this paper among them. Everything else redirects to the login screen. The list is not a pattern and carries no wildcard: a route is reachable anonymously only if it appears there literally, and a drift guard walks the route group in both directions, failing the build when a page is added without being declared or a declaration outlives its page.
  • Signed sessions. The cookie carries an HMAC-SHA256 token over the resolved identity — user, role, organization — with issue and expiry stamps, verified at the edge and in server components; it is valid for eight hours and revocable through a Session row before then.
  • Credentials without email. A username plus an access key in Crockford base32, stored only as a SHA-256 hash, compared in constant time, with uniform denial and exponential lockout. The key is minted at claim time and revealed once, so the plaintext never exists at rest in any form.
  • One tenant guard. The proxy strips any client-supplied tenant header and injects the verified one; routes call one helper that returns 403 when there is no tenant context, constrain reads to the tenant, answer 404 rather than 403 on cross-tenant existence so nothing leaks, and stamp ownership on writes so a client can never set it.
  • External keys and metering. API keys with a recognizable prefix are rate-limited per hashed key bucket; the metering seam records billable work per organization per UTC day after the response is computed and is never awaited.
Routes importing the tenant guard
457
repo@259feb2· 2026-11-21
Models carrying orgId
216
repo@259feb2· 2026-11-21
Routes calling meter()
7
repo@259feb2· 2026-11-21
Routes using the cache directive
7
repo@259feb2· 2026-11-21

Those four tiles are the honest state of adoption. The guard, the meter and the cache directive are each a few lines per route; the work that remains is mechanical and is tracked in §13 rather than assumed here.

§ 10

Every noun is a URL

A country, an event, a chokepoint, a facility, a sanctions program, a territory, a treaty: each is an addressable object with one metric contract, so comparison, linking and citation cost nothing per kind.

The object spine registers 7 kinds with canonical ids — ISO 3166-1 alpha-2 for countries, UN/LOCODE where one exists for facilities, slugs elsewhere — and every kind emits the same dossier shape: a list of metrics, each with a value, a unit, a direction where one is defensible, and its own provenance. The universal side-by-side zips two dossiers by metric key. It computes a delta only when both sides resolved a number — one side missing is reported as no comparison, never as a delta against zero — and it names a leader only when the metric declares a direction, because without one bigger is just bigger and calling it better would be an editorial claim the data does not support.

The object index is honest in a specific way: it lists what the warehouse can actually serve, not what a registry declares. A chokepoint with no port snapshot and no AIS box is still listed, and its count says zero — that is the coverage story, and hiding it would turn the index into a brochure. Events additionally carry a source race: who reported first, and by how many minutes.

Object kinds
7
repo@259feb2· 2026-11-21
Desks in the catalog
86
26 live · 55 building · 5 planned
repo@259feb2· 2026-11-21
Deep routes under desks
503
repo@259feb2· 2026-11-21
Page routes
390
repo@259feb2· 2026-11-21

The catalog status is itself a doctrine: live means built and wired to real data, building means partial real backing exists, and planned means designed and rendering an honest scaffold — never fabricated data. The split is measured from the catalog, not from a slide.

§ 11

Data strategy

Latency costs money. Breadth is free. The moat is the join, and the join is free.

There are two businesses a platform like this could be. One competes on latency — the equity consolidated tape, the options feed, a low-latency newswire — against firms with nine-figure data budgets, on data everyone else also buys. The other competes on correlation across domains that nobody joins in one place: public filings, government open data, disclosure feeds, sanctions regimes, humanitarian trackers, weather, orbit. Its floor cost is close to zero and nobody is contesting it, because the work is unglamorous integration rather than a rack in a data center. OpsHiyma chose the second, and the connector fleet is that choice made concrete.

Connectors in the tree
290
1 hand-written, the rest manifest-declared
repo@259feb2· 2026-11-21
Registered in the loader
288
a connector the loader does not name never runs
repo@259feb2· 2026-11-21
Need no key at all
224
78% of the 289 manifests
repo@259feb2· 2026-11-21
Domain classes
42
macro, sanctions, conflict, markets, space, …
repo@259feb2· 2026-11-21

Those are three different sets and the paper keeps them apart: 290 connector directories exist, 289 of them declare a manifest (1 is a hand-written client), and 288 are named in the loader registry — a connector the registry does not name is code that never runs, which is the kind of gap a single blurred number hides.

Histogram over every manifest’s declared freshness_target at measurement time; each example is a connector that declares that band
Freshness targetConnectorsExamples
realtime13GDELT, USGS seismic, NWS alerts, ISS position
near_realtime24Equity tape prints, Kalshi, NASA FIRMS, RSS wires
hourly69OpenSky, AIS, fusion crowd sweep
daily118OFAC, FINRA short volume, FRED, IDMC displacement, HDX
weekly51CFTC positioning, Statistics Canada
other14cadences outside the standard set (15min, fast)

The join

The end product is a terminal where every number traces to a public record, and its core loop runs entity → ownership → exposure → event → positioning: who is this thing (legal-entity and security-master identifiers), who controls it and are they sanctioned (13D/G and 13F filings against six sanctions regimes), what is it physically dependent on (trade, awards, grid, shipping and flights), what just happened to it (events, filings, rulemaking, disasters), and how is money placed against it (insider and institutional filings, short interest, futures positioning, crowd markets). Every incumbent holds every one of those datasets and cannot ask the question in one place, because they live in different products.

What is deliberately not built

Real-time options flow from the consolidated feed, a co-located equity execution terminal, low-latency newswire reaction, live sports odds from a paid feed, tasked satellite imagery. Each is a latency product. Where the platform touches options and equities today it does so on research and paper-trading lanes, and declining the latency race is the strategy rather than a limitation to apologize for in the interface.

§ 12

Evidence

Two kinds of figures. The first set was measured from the repository at build time and sealed; the second is read from the warehouse when you load this page, inside its own provenance envelope.

Measured from the code

Hover any chip for the exact method. All of these come from lib/whitepaper/facts.json, written by npm run whitepaper:facts at commit 259feb2 on 2026-11-21.

API routes
1,458
repo@259feb2· 2026-11-21
Pages
390
repo@259feb2· 2026-11-21
Routes with provenance
971
67% of routes
repo@259feb2· 2026-11-21
Routes honoring as-of
9
repo@259feb2· 2026-11-21
Source files
14,383
repo@259feb2· 2026-11-21
Lines of TypeScript
2,457,686
repo@259feb2· 2026-11-21
Modules under lib/
268
repo@259feb2· 2026-11-21
Unit tests
19,853
2371 files
repo@259feb2· 2026-11-21

Read from the warehouse now

Large-table counts are the planner’s live-tuple estimates (marked ≈), because an exact count over a multi-million-row table is a sequential scan and this is a page, not a job; governance tables are counted exactly. If the warehouse is unreachable when this renders, this block says so instead of remembering.

loading the warehouse…

The seal

datasetopshiyma-whitepaper@259feb2algorithmsha256-merkle, RFC 6962-style domain separationclaims38 figures, each { key, value, method }issued2026-11-21T15:28:18.321Z (the measurement time — the seal is a pure function of the facts)merkle rootddb9f461c574ebe70be8cdb9032be98bb46b0159ebea3b2bdfe9559a11651016

The full certificate with a per-figure inclusion proof is served at /api/whitepaper/facts.

{ key: 'apiRoutes', … }leaf = sha256(00 ‖ sha256(json)){ key: 'connectorManifests', … }leaf = sha256(00 ‖ sha256(json)){ key: 'prismaModels', … }leaf = sha256(00 ‖ sha256(json)){ key: 'testCases', … }leaf = sha256(00 ‖ sha256(json))node = sha256(01 ‖ left ‖ right)node = sha256(01 ‖ left ‖ right)merkleRoot — printed on this page
Each figure becomes a claim {key, value, method}; its sha256 is prefixed 0x00 as a leaf, pairs are prefixed 0x01 as nodes, and the root is printed on this page. Any figure can be proven against the root with its sibling path; changing one figure changes the root.
§ 13

What is not yet true

A paper about honesty that hid its own gaps would refute itself. Each row is a claim the platform wants to make, its state today, and the measured evidence.

Every served figure carries a provenance envelope.
PARTIAL
971 of 1458 route handlers import the envelope (67%). The rest return shapes that predate the doctrine or are pure control-plane endpoints.
No served path fabricates a value.
PARTIAL
The fabrication ratchet still tolerates 0 legacy offenses across 0 files; 20 served-path files contain Math.random() (some are legitimate ids and jitter). The ledger can only go down.
No vendor is awaited on a user request.
TRUE TODAY
The request-path ledger lists 0 files and 0 tolerated offenses. Ingest-side workers are exempted by path, not by comment.
Tenant-owned reads are isolated per organization.
PARTIAL
457 routes import the tenant guard and 216 of 509 models carry an orgId column. Global-warehouse reads need no guard; this paper does not claim every tenant-owned route has one.
Every billable call is metered.
NOT YET
The UsageMeter writer exists and 7 routes call it. Metering is a two-line adoption per route; adoption is the open work.
The terminal can be rewound to a past moment.
NOT YET
9 route honors an as-of parameter. Observations already store reportedAt beside ts, so the data can be replayed; the serving layer cannot yet.
The Trust Layer certifies the serving path.
PARTIAL
162 modules outside the library import it (the admin audit page, its API, the Truth Board leg-trust). Adapters exist for in-memory profiles only; the Postgres adapter is the next build.
Model outputs are compared semantically.
NOT YET
The consensus engine measures agreement with token-overlap (Jaccard) similarity and says so in code. Embeddings are the planned replacement.
The epistemic write path is populated.
PARTIAL
The Evidence writer and the Assessment desk exist and refuse correctly; the live counts in §12 show how few rows they hold today.
An independent party has certified the platform.
NOT YET
No third-party security or compliance certification is claimed anywhere in this paper. The gates described here are self-imposed and reproducible, which is a different thing.

Two further caveats. The desk catalog’s status split — 26 live, 55 building, 5 planned — is the platform’s own assessment against its own definitions, and “live” means wired to real data, not that every panel on that desk carries an envelope. And the deployment workflow defines an hourly ingestion job; whether that job is running at any given moment is something the live proof in §12 shows through the last successful run, not something this paper asserts.

§ 14

Roadmap

Convergence before width: close the foundation, prove one desk end-to-end with every cross-cutting guarantee, then fan every workstream onto that one spine.

PhaseWhat changesGate to pass
E1 · Promote the atomEvery ingested datum becomes a versioned, evidenced claim; the six-state status grows into the full epistemic class taxonomy, stamped at write timeClaim + Evidence become the central write path
E2 · Reconcile sources, not just modelsRe-point the consensus engine at multi-source data: corroboration, contradiction, outliers, confidence as a reusable service; persist the provenance chainReconciliation API on real vendor disagreement
E3 · Belief-timeA third time axis — when we came to believe a value and how confidence moved — and a query surface that filters on epistemics and answers AS OF BELIEFAs-of served from DeskEnvelope first, then everywhere
E4 · Transparency logContent-address claims and evidence into a durable Merkle log; export proofs for compliance and data marketplacesA /verify endpoint over persisted certificates
E5 · Productize the substrateThe Trust Layer with a Postgres adapter pointed at this warehouse as the lighthouse reference, then external estatesA published Trust Audit of the platform’s own warehouse

In parallel, the mechanical adoption that §13 names — the tenant guard, the meter, the cache directive and the envelope on every remaining route — proceeds under the existing gates, and embeddings replace token overlap across consensus, fusion and clustering. None of it widens the plan; each item is an instance of the pattern the rest of this paper describes.

§ 15

Verify this paper

Do not take the figures on trust. Re-measure the tree, run the gates, and check any figure against the root.

From a checkout of the repository
npm run whitepaper:facts      # re-measure the tree → lib/whitepaper/facts.json
npm test                      # includes tests/whitepaper-facts.test.ts: drift tolerance + seal
npm run lint:fabrication      # the fabrication ratchet
npm run lint:warehouse-path   # the request-path ratchet
npm run lint:series-keys      # the warehouse key contract (needs a reachable warehouse)
npm run trust:audit           # the Trust Layer pipeline on its demo dataset
The seal, from a signed-in session
GET /api/whitepaper/facts
→ { repo, seal: { certificate, claims[ { key, value, method, contentHash, proof[] } ] }, warehouse, provenance }

leaf  = sha256("00" + sha256(stableStringify({ key, value, method })))
node  = sha256("01" + left + right)      # odd node pairs with itself
root  = fold leaves pairwise until one hash remains — it must equal certificate.merkleRoot
Check one figure by hand (Node, from the repository root)
import { hashClaim, verifyProof } from './lib/trust-layer/certificate';
const res = await fetch('/api/whitepaper/facts', { headers: { cookie } }).then((r) => r.json());
const c = res.seal.claims.find((x) => x.key === 'apiRoutes');
const leaf = hashClaim({ key: c.key, value: c.value, method: c.method });
console.log(leaf === c.contentHash, verifyProof(leaf, c.proof, res.seal.certificate.merkleRoot));

If a re-measurement disagrees with a printed figure by more than the drift tolerance, the unit test fails and the fix is to regenerate the snapshot — there is no way to edit a figure by hand that survives the test.

§ 16

Glossary & references

Provenance envelope
The record attached to every served value: source, asOf, fetchedAt, one of six statuses, optional confidence, a modeled flag and a note.
Serving envelope
A dated, schema-versioned JSON product a heavy desk stores and returns instead of recomputing; reassembled from canonical atoms after the response, for today only.
Canonical atom
Event, Series/Observation, GeoFeature, Entity/Relationship, Claim, Evidence — the warehouse rows a desk is allowed to read.
Ratchet
A CI gate whose ledger of tolerated legacy offenses can only decrease; more fails, fewer fails until ratcheted, unknown files start clean.
Drift guard
A test that walks the real tree and compares it with a registry or snapshot, so coverage is a property of the data.
Verifiable Claim
The target atom: subject–predicate–object, value or distribution, epistemic class, evidence chain, corroboration and contradiction, calibrated confidence, proof.
ACH
Analysis of Competing Hypotheses. Hypotheses are ranked by weighted inconsistency only; consistency never counts; under-filled matrices refuse to rank.
Admiralty code
Source grading: reliability A–F, credibility 1–6. The basis of a grade — authored, derived, earned, unrated — is always shown.
ICD 203
The intelligence-community analytic standard. Confidence (low / moderate / high) has no probability; likelihood uses the seven-band yardstick.
Brier score
Σ (pᵢ − oᵢ)² over outcomes; 0 is perfect, 2 maximally wrong. A skill score against a baseline says whether the desk added anything.
Nowcast
A composite index whose confidence is the coverage of its applicable inputs, and which reports NO_DATA rather than a default when they are missing.
Merkle seal
A SHA-256 Merkle root over a set of claims with leaf/node domain separation and per-claim inclusion proofs.
Tenant guard
One helper that binds a request to its verified organization and constrains every tenant-owned read and write to it.
Tool loop
The bounded model-calls-tools cycle behind the copilot, restricted to tools that require no approval.

Sources in the repository

  • lib/provenance.tsThe envelope and the doctrine sentence.
  • docs/warehouse-request-path.mdThe request-path freeze and the serving-envelope contract.
  • scripts/check-fabrication.mjs · scripts/check-warehouse-path.mjs · scripts/check-series-keys.mjsThe three honesty gates and their ledgers.
  • prisma/schema.prismaThe warehouse spine, including the bitemporal Observation and the content-hashed Evidence.
  • lib/intel/ach-kernel.ts · lib/intel/evidence-writer.ts · lib/intel/types.tsThe ACH kernel, the Evidence writer’s refusals, and the ICD 203 vocabulary.
  • lib/political-risk/ledger-pure.tsThe election call ledger and its Brier scoring.
  • lib/fusion/divergence.ts · lib/escalation/index-pure.tsStandardized-leg divergence and the coverage-confident nowcast.
  • lib/trust-layer/Audit, trust score, reconciliation, contamination, certificate.
  • lib/copilot/tool-loop.ts · lib/copilot/tool-bridge.ts · lib/sif/ace/The tool loop, its safety gate, and the consensus engine.
  • lib/auth/org-scope.ts · lib/auth/session.ts · lib/auth/public-routes.ts · proxy.tsThe tenant guard, signed sessions, the page gate and the default-deny proxy.
  • lib/objects/The object spine: kinds, dossiers, the universal compare.
  • docs/epistemic-warehouse-thesis.md · docs/trust-layer-gtm.md · docs/zero-cost-data-strategy.md · docs/completion-program.mdThe thesis, the commercialization of the Trust Layer, the data strategy, and the convergence program.
  • lib/whitepaper/This paper’s collector, snapshot, seal and warehouse proof.

Questions about the platform go through support. This document is a public draft and will be re-sealed whenever the snapshot is regenerated.

Hiyma

Operational intelligence with its provenance attached. Desks read the warehouse. Empty is empty.

Network
DocsStatusSecurity
Legal
PrivacyTermsExportCompliance
Modules
SignalsArmsMarketsSatellite
Uplink
Use CasesPricingAPIEnterpriseSupport
System status
© 2026 Hiyma · closed beta · access by approval
last successful ingest run 2026-11-21 14:07 ET
Series observations
≈10,015,591
time-stamped values, unique on (series, ts)
Warehouse series
≈282,962
unique on (provider, dataset, seriesKey)
Canonical events
≈1,070,564
ingested world events with source + occurredAt
Geo features
≈296,574
point / polygon / hex layers with validFrom
Raw records
≈1,309,433
payload-hashed originals behind every atom
Entities
≈131,592
resolved people, orgs, vessels, countries
Sanction designations
≈686
across the wired sanctions regimes
Maritime positions
≈172,780
AIS fixes held by the logistics worker
Headlines
≈536,896
labelled once at ingest
Ingest runs
≈219,431
one audit row per connector run
Registered sources
334
Source rows (one per connector code)
Sources with a successful run
281
distinct sourceId with an IngestRun status = success
Successful ingest runs
208,022
IngestRun status = success
Failed ingest runs
11,409
IngestRun status = error — kept, never hidden
Serving envelopes
12,002
DeskEnvelope rows (dated desk products)
Claims
29,661
subject–predicate–object claims
Evidence rows
5
content-hashed citations (the epistemic write path is young)
Assessments
1
ACH assessments on the Assessment Desk
Audit log rows
608
governance trail incl. control-plane gates

Large-table counts are planner estimates (pg_stat_user_tables.n_live_tup); governance tables are exact counts.

≈22.6M
warehouse rows
11
intelligence domains
114 / 590
tables holding data
188
countries tagged, 7 days
288
registered connectors
509
warehouse models
1,458
API routes
19,853
unit tests

Top row read from the warehouse at render; bottom row measured from the repository and sealed.

hiyma console
warehouse 14:11 ET
HIYMA console · public tier · aggregates only
22,626,027 rows · 11 domains · 114 of 590 tables hold data
Type HELP, or try CENSUS · MAP · WIRE · EVENTS · DARK · CO IR
Figure 5The aggregates behind §12, queryable. Every answer is computed from the snapshot this page was rendered with, in your browser, with no request leaving it — so the figures the paper prints can be cross-checked against the warehouse without trusting the paper.
The warehouse, by area
11 domains · ≈22,626,027 rows · 114 of 590 tables hold data
Macro & reference series
10.3M
57% · 2/2 tables
Equities & tape
2.7M
15% · 5/7 tables
Provenance & audit
1.6M
9% · 3/4 tables
World events
1.4M
8% · 3/3 tables
Institutional filings
873K
5% · 7/7 tables
News & wire
541K
3% · 2/2 tables
Geospatial
Cross-domain indices
Maritime & mobility

Area is row count. These 11 domains account for 80.52% of every row in the warehouse — the rest is bookkeeping.

Figure 2The warehouse by area, read at the moment this page rendered. Each tile is a set of real Prisma models named in prisma/schema.prisma; area is row count. The spread runs four orders of magnitude, which is why this is a treemap and not a bar chart.
Where the wire pointed · last 7 days
188 countries · 55,706 headlines
United States — 3,763 tagged headlinesFrance — 2,287 tagged headlinesChina — 1,140 tagged headlinesCanada — 803 tagged headlinesRussia — 519 tagged headlinesUkraine — 439 tagged headlinesUnited Kingdom — 367 tagged headlinesGermany — 338 tagged headlinesIran — 306 tagged headlinesIndia — 282 tagged headlinesIsrael — 247 tagged headlinesJapan — 230 tagged headlinesCentral African Republic — 199 tagged headlinesAustralia — 186 tagged headlinesPalestine — 180 tagged headlinesItaly — 175 tagged headlinesSingapore — 123 tagged headlinesTaiwan — 116 tagged headlinesSpain — 110 tagged headlinesRéunion — 108 tagged headlinesBrazil — 94 tagged headlinesMexico — 93 tagged headlinesPhilippines — 93 tagged headlinesSaudi Arabia — 89 tagged headlinesSouth Korea — 82 tagged headlinesHong Kong — 80 tagged headlinesSwitzerland — 72 tagged headlinesUnited Arab Emirates — 72 tagged headlinesSweden — 71 tagged headlinesGreece — 66 tagged headlinesIreland — 65 tagged headlinesTurkiye — 64 tagged headlinesYemen — 61 tagged headlinesSouth Africa — 60 tagged headlinesRwanda — 59 tagged headlinesMyanmar — 57 tagged headlinesNetherlands — 54 tagged headlinesVietnam — 50 tagged headlinesThailand — 48 tagged headlinesArgentina — 46 tagged headlinesPoland — 46 tagged headlinesNorway — 46 tagged headlinesDenmark — 45 tagged headlinesPakistan — 45 tagged headlinesMalaysia — 44 tagged headlinesIndonesia — 44 tagged headlinesBelgium — 42 tagged headlinesPortugal — 37 tagged headlinesCzechia — 35 tagged headlinesLebanon — 35 tagged headlinesBahrain — 35 tagged headlinesHaiti — 34 tagged headlinesNew Zealand — 33 tagged headlinesSenegal — 33 tagged headlinesVenezuela — 32 tagged headlinesPuerto Rico — 31 tagged headlinesIraq — 31 tagged headlinesFinland — 30 tagged headlinesBangladesh — 29 tagged headlinesNigeria — 27 tagged headlinesEgypt — 25 tagged headlinesPeru — 23 tagged headlinesCambodia — 23 tagged headlinesEstonia — 23 tagged headlinesColombia — 21 tagged headlinesQatar — 19 tagged headlinesAustria — 19 tagged headlinesNorth Korea — 19 tagged headlinesKenya — 19 tagged headlinesGreenland — 18 tagged headlinesHungary — 16 tagged headlinesUzbekistan — 15 tagged headlinesDR Congo — 15 tagged headlinesMacao — 13 tagged headlinesSri Lanka — 13 tagged headlinesSyria — 13 tagged headlinesMayotte — 12 tagged headlinesBermuda — 12 tagged headlinesBulgaria — 12 tagged headlinesCosta Rica — 12 tagged headlinesGuyana — 12 tagged headlinesRomania — 12 tagged headlinesMozambique — 12 tagged headlinesCroatia — 11 tagged headlinesMali — 11 tagged headlinesCôte d’Ivoire — 11 tagged headlinesJersey — 11 tagged headlinesMorocco — 10 tagged headlinesLuxembourg — 10 tagged headlinesKuwait — 10 tagged headlinesEl Salvador — 10 tagged headlinesNepal — 10 tagged headlinesSerbia — 9 tagged headlinesGuadeloupe — 9 tagged headlinesOman — 9 tagged headlinesMalawi — 9 tagged headlinesAfghanistan — 8 tagged headlinesCyprus — 8 tagged headlinesSomalia — 8 tagged headlinesTogo — 7 tagged headlinesGhana — 7 tagged headlinesKazakhstan — 7 tagged headlinesEthiopia — 7 tagged headlinesSudan — 7 tagged headlinesMartinique — 7 tagged headlinesUruguay — 6 tagged headlinesEcuador — 6 tagged headlinesBenin — 6 tagged headlinesCayman Islands — 6 tagged headlinesLithuania — 6 tagged headlinesTanzania — 6 tagged headlinesBolivia — 6 tagged headlinesGuatemala — 6 tagged headlinesLibya — 5 tagged headlinesAntigua & Barbuda — 5 tagged headlinesUganda — 5 tagged headlinesSt. Pierre & Miquelon — 5 tagged headlinesKosovo — 5 tagged headlinesHonduras — 5 tagged headlinesBurkina Faso — 5 tagged headlinesCameroon — 5 tagged headlinesLatvia — 5 tagged headlinesSlovenia — 5 tagged headlinesJordan — 5 tagged headlinesAlbania — 5 tagged headlinesGambia — 4 tagged headlinesParaguay — 4 tagged headlinesAzerbaijan — 4 tagged headlinesAmerican Samoa — 4 tagged headlinesDominican Republic — 4 tagged headlinesAngola — 4 tagged headlinesZimbabwe — 4 tagged headlinesSlovakia — 4 tagged headlinesPanama — 4 tagged headlinesMonaco — 4 tagged headlinesGuinea — 4 tagged headlinesIceland — 4 tagged headlinesChile — 4 tagged headlinesBotswana — 3 tagged headlinesSt. Martin — 3 tagged headlinesLiechtenstein — 3 tagged headlinesSouth Sudan — 3 tagged headlinesTrinidad & Tobago — 3 tagged headlinesArmenia — 3 tagged headlinesBosnia & Herzegovina — 3 tagged headlinesMadagascar — 3 tagged headlinesCuba — 3 tagged headlinesZambia — 3 tagged headlinesMauritius — 3 tagged headlinesAlgeria — 3 tagged headlinesChristmas Island — 3 tagged headlinesCape Verde — 3 tagged headlinesSuriname — 3 tagged headlinesBritish Virgin Islands — 3 tagged headlinesBahamas — 3 tagged headlinesEritrea — 3 tagged headlinesNorth Macedonia — 2 tagged headlinesMaldives — 2 tagged headlinesBarbados — 2 tagged headlinesJamaica — 2 tagged headlinesGabon — 2 tagged headlinesMongolia — 2 tagged headlinesBelarus — 2 tagged headlinesBrunei — 2 tagged headlinesNicaragua — 2 tagged headlinesChad — 2 tagged headlinesVatican City — 2 tagged headlinesGeorgia — 2 tagged headlinesBhutan — 2 tagged headlinesAruba — 2 tagged headlinesSt. Kitts & Nevis — 2 tagged headlinesTajikistan — 1 tagged headlinesGuernsey — 1 tagged headlinesPapua New Guinea — 1 tagged headlinesSolomon Islands — 1 tagged headlinesUS Virgin Islands — 1 tagged headlinesTurkmenistan — 1 tagged headlinesFiji — 1 tagged headlinesMicronesia — 1 tagged headlinesSamoa — 1 tagged headlinesGrenada — 1 tagged headlinesSierra Leone — 1 tagged headlinesSt. Barthélemy — 1 tagged headlinesGibraltar — 1 tagged headlinesTunisia — 1 tagged headlinesBurundi — 1 tagged headlinesFrench Guiana — 1 tagged headlinesNamibia — 1 tagged headlines
🇺🇸United States3,763
🇫🇷France2,287
🇨🇳China1,140
🇨🇦Canada803
🇷🇺Russia519
🇺🇦Ukraine439
🇬🇧United Kingdom367
🇩🇪Germany338

A country with no dot carried no tagged headline this week. Nothing is shaded to imply coverage that is not there, and no boundary is drawn.

Figure 3Every country the headline labeller tagged in the last seven days, placed at the centroid the desks use and sized by volume. No borders are drawn and no region is shaded: a country with no dot carried no tagged headline, which is a statement about coverage rather than about the world.

Who is feeding the tape · 7 days

40 publishers
  1. 01SEC FORM 49,787
  2. 02BLUESKY8,883
  3. 03FISCAL.AI8,846
  4. 04PR NEWSWIRE3,910
  5. 05GURUFOCUS3,627
06SEEKING ALPHA
2,029
  • 07BFMTV1,970
  • 08MOTLEY FOOL1,524
  • 0924/7 WALL ST1,328
  • 10SEC 8-K1,189
  • 30 more held. Volume across 55,706 headlines. Counts only — what they published is behind the key.

    What kind of thing happened · 7 days

    57 categories
    1. 01News story41,700
    2. 02Market resolution21,600
    3. 03Satellite wildfire4,626
    4. 04Insider transaction3,832
    5. 05Borrow change3,648
    6. 06Seismic event2,275
    7. 07Institutional holdings2,044
    8. 08Corporate event1,958
    9. 09Issuer filing1,887
    10. 10Passive stake1,673

    47 more held. One table, one clock. A wildfire and an arms transfer are peers here, which is what makes them comparable.

    Sanctions regimes held

    24 regimes
    1. 01UK · OFSI11,790
    2. 02Russia-Eo140245,664
    3. 03EU · Ukr4,786
    4. 04UK · OFSI · Russia3,581
    5. 05Ca Sema Russia3,394
    6. 06Sdgt2,193
    7. 07Sdntk1,334
    8. 08Iran-Eo139021,043
    9. 09EU · Irn993
    10. 10EU · Blr731

    14 more held. Designations by issuing list. Keys are displayed readably; the warehouse keeps the publisher’s own spelling.

    Reference & macro providers

    110 providers
    1. 01SEC125,599
    2. 02FEED40,430
    3. 03UNHCR17,153
    4. 04WORLDBANK15,924
    5. 05FINRA14,750
    6. 06UN12,798
    7. 07IMF10,889
    8. 08kalshi9,564
    9. 09polymarket6,570
    10. 10MASSIVE5,340

    100 more held. Series per publisher. Institutional open data, most of it keyless.

    Figure 4What the rows are about, by dimension: who filed them, what kind of event arrived, which sanctions regimes are held and which institutions publish the reference series. Counts only — the public tier serves no row.
    Arrivals per hour · last 48 hours
    peak 2,746/h · 0 silent hours
    48 h ago → now, left to right, two rows
    none2,746/h
    Figure 1Rows reaching the warehouse per hour over the last 48 hours, counted on the arrival clock and zero-filled by Postgres so an empty hour is a measured zero rather than a missing bucket. Silences are visible as holes; the paper does not smooth them.