Skip to content

Worked example

The ten stages carry one worked example: Meridian Engineering, a fictional mid-size AEC firm. This page shows the result — the same slice expressed as live data across all four layers.

# Each name is followed by the Shape the Thing declares in `shapes`;
# no name segment selects a Shape.
# Layer 1 — grounding (per-source repo; compact names under its NamingContracts)
wh:pub/us.environmental.epa-npdes/dmr-records/nc/NC0071234/001/00610/2025-06 # shapes: [DmrRecord]
{ reportedValueNumeric: 7.3, unitCode: "19", statisticCode: "30",
sourceArtifact -> artifacts/epa-npdes-dmr@v7 }
artifacts/epa-npdes-dmr@v7 # shapes: [SourceArtifact]
{ source -> sources/epa-npdes-dmr, semanticSha256: "ab12...",
mediaType: "application/x-ndjson",
originalUrl: "https://echo.epa.gov/...",
durableUri: "<artifact-store>/semantic/sha256/ab12...",
acceptedAt: "2025-07-15T12:00:00Z",
canonicalizationPolicy: "epa-npdes-dmr-canonical-jsonl/v1" }
# Layer 2 — identity (curated repo, identity namespace)
wh:pub/us.infrastructure/identity/facilities/epa-frs/110012345678 # shapes: [Facility]
identity/relations/epa-npdes-NC0071234--epa-frs-110012345678 # shapes: [Arc]; named per predicate recipe
identity/facility-references/epa-npdes/NC0071234 # shapes: [RefersToFacility]
about identity/relations/epa-npdes-NC0071234--epa-frs-110012345678
{ resolutionMethod: "epa-frs-program-link", evidenceWrefs: [...] }
# Layer 3 — domain (same repo, water namespace)
water/measurements/nc/NC0071234/001/ammonia-nitrogen/2025-06 # shapes: [Measurement]
{ parameter: water/parameters/ammonia-nitrogen, value: 7.3,
unit: water/units/mg-per-l, statistic: water/statistics/monthly-average,
derivedFrom: <grounded record wref>, transformationPolicy: "dmr-norm/v3" }
water/performance-patterns/millbrook-ammonia-warm-season-2022-2025 # shapes: [PerformancePattern]
about identity/facilities/epa-frs/110012345678
{ patternClass: recurring-warm-season-elevation, methodVersion: "1.0", evidenceWrefs: [...] }
# Layer 4 — beliefs (a public prediction; a private decision composing it)
wh:pub/us.infrastructure/water/pressure-assessments/millbrook-2026 # shapes: [RegulatoryPressureAssessment]
{ summary: "public evidence indicates treatment-upgrade pressure within 24 months",
evaluationAsOf: ..., evidenceWrefs: [water/performance-patterns/..., ...] }
wh:meridian/intelligence/outreach/town-of-millbrook/2026Q3 # shapes: [WarrantsOutreach]
about wh:pub/us.infrastructure/identity/organizations/nc/town-of-millbrook
{ evidenceWrefs: [ icp-fit/..., capabilities/..., # private beliefs
wh:pub/us.infrastructure/water/pressure-assessments/millbrook-2026 ] }

A skeptical reader starts at the WarrantsOutreach assertion outreach/town-of-millbrook/2026Q3 and walks, wref by wref, to a pinned EPA semantic-snapshot version and hash on one side and Meridian’s own capability model on the other. That traversal existing, being cheap, and never being fabricated is what this whole method is for.