DEPLOYThe reference layer for physical AI

Documentation

JSON-LD deploy: namespace

DEPLOY's custom JSON-LD vocabulary at https://deploy.report/schema/v1/. Per-property reference for the 23 properties published in entity-page JSON-LD. The deploy: namespace is the channel for verification posture, autonomy- boundary classifications, brain-provider architectures, within-entity sagas, and funding-round graphs that schema.org alone cannot express.

Design

Namespace IRI

https://deploy.report/schema/v1/. JSON-LD consumers do not need the IRI to resolve to fetch the vocabulary document; the namespace identifies the deploy: prefix, and the vocabulary is documented here.

Properties

Methodology metadata

The full methodology JSON-LD block emits on every entity page with reviewStatus = reviewed. Superset of the framework_metadata REST block; carries everything framework_metadata does plus per-entity methodology framing (verification posture narrative, confidence tier rationale, source tier breakdown).

deploy:methodology

Container for the methodology block. Composes verification posture, confidence tier, source tier breakdown, cap-flags, architectural position, and within-cohort pair membership into one structured surface.

Range: deploy:MethodologyMetadata

Emitted on: Every reviewed entity page across all 7 entity types (model / company / brain / deployment / incident / regulation / location).

Worked example: ringconn-gen2 (Model)

Related surface: v2.get_methodology MCP tool; GET /v1/methodology/{slug}

deploy:verificationStatus

verified / claimed-not-verified / claimed-not-cleared / aggregator-drift-detected / unreviewed. The top-level verification posture for the entity.

Range: string

Emitted on: Inside deploy:methodology block.

deploy:confidenceTier

High / medium / low / cap-flag-required. Set by the framework metadata builder based on source quality, reviewStatus, and anchored drift patterns.

Range: string

Emitted on: Inside deploy:methodology block.

deploy:capFlags

Anchored drift patterns that require honest framing. Each cap_flag carries claim, honest_status, common_drift, and drift_pattern_slug. The framework refuses to drop a record's cap-flags for cleaner data; they are part of the record's verified state.

Range: array of deploy:CapFlag

Emitted on: Inside deploy:methodology block (and the standalone framework_metadata in REST).

Worked example: ringconn-ahi-fda-conflation cap_flag on the RingConn Gen 2 methodology

Related surface: v2.analyze_aggregator_drift MCP tool

deploy:architecturalPosition

Cohort + sub-cohorts the entity participates in. Cohort is the form_factor; sub_cohorts are the architectural-axis pair memberships the entity is anchored to.

Range: deploy:ArchitecturalPosition

Emitted on: Inside deploy:methodology and framework_metadata.

deploy:withinCohortVerifiedVsClaimedPair

Canonical worked example pair the entity participates in (when applicable). The pair anchors a verification-posture axis (FDA-clearance, AI-substance-gradient, business-model variance, etc.) to two or three concrete registry entities; pair_with carries the counterparts.

Range: deploy:PairMembership

Emitted on: Inside deploy:methodology and framework_metadata.

Related surface: v2.list_canonical_worked_examples MCP tool

deploy:verificationDepth

sources_count + primary_source_types array (the source quality classifier output). Surfaces how the verification posture was earned.

Range: deploy:VerificationDepth

Emitted on: Inside deploy:methodology and framework_metadata.

deploy:frameworkSchemaVersion

Tracks FRAMEWORK_SCHEMA_VERSION. Bumped when the shape of framework_metadata changes. Lets downstream consumers branch on version cleanly.

Range: string

Emitted on: Inside deploy:methodology and framework_metadata.

Autonomy boundary classification

Four-tier classification of where autonomy lives in the operator <-> machine loop. Emits on Model and Deployment entity pages where the underlying model has been classified into the SSoT.

deploy:autonomyBoundary

Container for autonomy-boundary classification: tier (autonomous-execution / ai-augmented-operator-controlled / replacement-robotics-teleoperated / assistive-co-pilot) and per-entity rationale.

Range: deploy:AutonomyBoundaryClassification

Emitted on: Model entity page; Deployment entity page (inherits via model slug).

Worked example: stryker-mako (ai-augmented-operator-controlled tier)

Related surface: v2.classify_autonomy_boundary MCP tool; GET /v1/taxonomies/autonomy-boundary

deploy:tier

Full tier definition: slug, label, definition, verification criteria, common drift patterns, cohort applicability, editorial reference.

Range: deploy:AutonomyBoundaryTier

Emitted on: Inside deploy:autonomyBoundary.

deploy:entityRationale

The per-entity rationale for the tier assignment; null when the entity is not a canonical example for the tier (still classified, no specific rationale).

Range: string

Emitted on: Inside deploy:autonomyBoundary (and parallel structures).

Brain provider architecture

Four-way classification of brain provider architecture (captive vs third-party vs hybrid vs research-only). Emits on Brain entity pages where the brain has been classified.

deploy:brainProviderArchitecture

Container for brain-provider-architecture classification: architecture (captive-vertically-integrated / third-party-foundation-model / hybrid-licensed / research-only) and per-entity rationale. Affects counterparty risk, integration patterns, interoperability posture.

Range: deploy:BrainProviderArchitectureClassification

Emitted on: Brain entity page.

Worked example: 1x-redwood (captive-vertically-integrated)

Related surface: v2.classify_brain_provider_architecture MCP tool; GET /v1/taxonomies/brain-provider-architecture

deploy:architecture

Full architecture definition: slug, label, definition, verification criteria, common drift patterns, editorial reference.

Range: deploy:BrainProviderArchitectureDefinition

Emitted on: Inside deploy:brainProviderArchitecture.

Within-entity sagas and feature references

SAME-entity feature-scope verification posture variance. Distinct from canonical worked example pairs (which operate cross-entity). Emits on Model and Brain pages where sagas or feature references are registered.

deploy:withinEntitySagas

Verification posture variance over time or across jurisdictions for a single feature on the entity. Each saga carries timeline (event by event) plus jurisdictional variance entries (e.g., Apple Watch SpO2 availability by region).

Range: array of deploy:WithinEntitySaga

Emitted on: Model or Brain entity page when the entity has registered sagas.

Worked example: apple-watch-spo2-availability-saga on the apple-watch model

Related surface: v2.get_within_entity_saga MCP tool; GET /v1/sagas/{entity_slug}

deploy:withinEntityFeatureReferences

Single-feature feature-scope verification posture (manufacturer position vs authoritative position). Used when one feature on the entity carries materially different posture from the entity's other features (Whoop BPI vs other Whoop features).

Range: array of deploy:WithinEntityFeatureReference

Emitted on: Same entities as deploy:withinEntitySagas.

Worked example: whoop-bp-insights-feature-reference on the whoop model

Related surface: v2.list_within_entity_feature_references MCP tool; GET /v1/sagas/{entity_slug}

Media attribution

Verification-posture signal on media assets. Lets AI engines + crawlers branch on whether a VideoObject / ImageObject is independent-named-site evidence or manufacturer-supplied product-showcase. Emitted via additionalProperty PropertyValue with the deploy: namespace propertyID.

deploy:mediaJoinType

Surfaces the verification posture of the media asset structurally. 'deployment-linked' = asset carries a MediaAssetDeployment join and surfaces in the Deployment-verified media section; 'model-linked' = asset is manufacturer-attributed at the model level with no deployment context, surfaces in the Manufacturer-attributed media section. Carried via schema.org PropertyValue with propertyID = https://deploy.report/schema/v1/mediaJoinType.

Range: string (deployment-linked | model-linked)

Emitted on: Every ImageObject and VideoObject on Model entity pages when the page render passes a joinType through to buildMediaObjectJsonLd.

Worked example: figure-02 model page: Spartanburg BMW deployment-verified video carries deployment-linked; product-render assets carry model-linked.

People graph

Employment, founding events, and board seats linked to Company entities. Each block honest-absence (omitted) when the company has no edges of that type. Auto-emits as Agent A's editorial backfill populates the people-graph tables.

deploy:people

Full employment edge graph: role, role_title, startDate, endDate, currentRole flag, role_basis (verification depth), per-person identity inline.

Range: array of deploy:PersonCompany

Emitted on: Company entity page when at least one PersonCompany edge exists.

Worked example: apptronik (Jeff Cardenas Co-founder & CEO + co with verified_secondary role_basis)

Related surface: GET /v1/companies/{id}?include=people; GET /v1/companies/{id}/people; v2.list_people_by_company MCP tool

deploy:founders

Founding events with cofounderStatus + stillAtCompany signal. Distinct from PersonCompany.role='founder' (the role surface); this is the founding-event surface.

Range: array of deploy:PersonFounder

Emitted on: Company entity page when at least one PersonFounder edge exists.

Related surface: GET /v1/companies/{id}?include=founders; v2.get_founders MCP tool

deploy:board

Board seats with role (chair / director / observer / advisor), startDate, endDate, current flag.

Range: array of deploy:PersonBoard

Emitted on: Company entity page when at least one PersonBoard edge exists.

Related surface: GET /v1/companies/{id}?include=board; v2.get_board MCP tool

deploy:person

Person identity reference: slug + name. The /people/[slug] entity page surface is a follow-up; the inline reference makes the graph traversable from the company-page JSON-LD today.

Range: deploy:Person

Emitted on: Inside each edge block (deploy:people / deploy:founders / deploy:board).

Related surface: GET /v1/people/{slugOrId}; v2.get_person MCP tool

Investor graph

Funding rounds and investor participations linked to Company entities. Honest-absence (block omitted) when the company has no rounds on record. Auto-emits as Agent A's editorial backfill populates rounds.

deploy:fundingRounds

Funding rounds for the company. Each round carries roundName, announcedAt, amountUsd, postMoneyValuationUsd, amountBasis (exact / reported / undisclosed), and a participations array with per-investor identity inline.

Range: array of deploy:FundingRound

Emitted on: Company entity page when the company has at least one funding round on record.

Worked example: apptronik (Company; Series A + Series A extension rounds)

Related surface: GET /v1/companies/{id}?include=funding-rounds; GET /v1/funding-rounds/{id}

deploy:participations

Per-investor participation in a round. Carries role (lead / co_lead / follow / strategic / participant) and per-investor identity shorthand (slug, name, kind).

Range: array of deploy:InvestorParticipation

Emitted on: Inside each deploy:FundingRound.

deploy:investor

Investor identity reference. Carries slug, name, kind (venture / growth / strategic_corporate / sovereign_wealth / angel / accelerator / government / other). Full Investor entity page surface is forthcoming.

Range: deploy:Investor

Emitted on: Inside deploy:participations; will emit on the future /investors/{slug} entity page (builder ready in v02_jsonld).

Related surface: GET /v1/investors/{slugOrId}

See also