An official website of the Disclosure Foundation
RecordsObservablesOriginsClaimsScoringOnboardDocs

Records

The unit of record

Observables

Detection

Origins

Classification

Claims

Attribution

Scoring

Evaluation

Onboard

Bring your dataset

Docs

Implement the standard

GitHub

Open standards and tools for UAP research.

RecordsObservablesOriginsClaimsScoringOnboardDocsGitHubDonate

© 2026 Disclosure Foundation. All rights reserved.

The Standard · Part 4 of 5 — Attribution

Claims

Every assessment of a case is recorded as a claim: who made it, when, why, and the evidence behind it. When evaluators disagree, competing claims sit side by side instead of overwriting each other.

@disclosureos/recordsDocumentation

What every claim carries

An attributed, evidence-backed assertion (Attribution + the evidence that justifies it). A record holds facts; a claim is how someone’s assessment attaches to those facts without becoming one of them. Every claim carries the same structure: who made it, when, their reasoning, and pointers to the evidence that backs it. No claim is ever treated as the official answer — deciding which claims to trust comes later, in scoring.

  • evaluatedAt

    When this judgement was made (ISO 8601 datetime).

  • evaluatedBy

    Who made this judgement — a person, agency, or agent identifier.

  • rationale

    Why this judgement was reached — the evaluator's stated reasoning.

  • evidenceRefs

    Refs to in-record evidence justifying this claim, e.g. "media:<id>", "sensor:<id>".

Evidence types

A claim points at the evidence that justifies it using a typed pointer into the record itself — four kinds, written as kind:id. That pointer is checked against evidence already on the record, so a claim can never cite something that is not there.

  • media:<id>

    A media attachment on the record (photo, video, audio, document).

  • sensor:<id>

    A sensor reading in the record’s sensor evidence.

  • physical:<id>

    A physical evidence item (forensic extension).

  • testimony:<id>

    A testimony statement (forensic extension).

Two axes of uncertainty

UAP data is contested by nature, so the model keeps two kinds of disagreement separate and never collapses one into the other. Within a single claim, one evaluator can hedge. Across claims, different evaluators can flatly disagree — and both claims stay on the record.

Across claims — disagreement

Claims come as a list

A record holds a list of claims, not a single answer. Competing claims from different evaluators coexist; none is overwritten or marked as the official one.

USS Nimitz “Tic Tac” (2004) · Instantaneous Acceleration

  • Navy aviator (firsthand)Measured85%

    Radar tracked descent from ~80,000 ft to sea level in seconds.

  • AARO (later review)Not Indicated60%

    Available sensor data judged insufficient to establish the maneuver.

Within a claim — hedging

Primary plus alternatives

A single origin claim names one primary hypothesis and can weighalternativeHypothesesalongside it — one evaluator expressing graded uncertainty.

Independent analyst · one claim

1.1.3Extraterrestrialprimary40%
1.1.1.2.1Governmentalalternative25%

Two first-party claim shapes

Observable claims and origin claims share the same core fields and add only the few fields that make them their kind. Each attaches to its own dedicated place on the record.

Observable Claim

@disclosureos/observables

An attributed observable claim, keyed by its map position (no observableId).

Adds

  • level

    Evidentiary tier of an observable assessment — how strongly the signal is established.

  • confidence

    Confidence in the assessment, in [0,1]. 0 = none, 1 = absolute.

Shared fields

  • evaluatedAt
  • evaluatedBy
  • rationale
  • evidenceRefs

Place on the record

Observable assessments

A list of observable claims for each Technology or Biologics signal.

Code: observableAssessments (ObservableClaim[])

Per observable id, a list of claims about how strong the evidence for that signal is.

Origin Claim

@disclosureos/origins

An attributed origin-hypothesis classification for an observation.

Adds

  • primaryHypothesis

    OCS node id of the primary hypothesis.

  • confidence

    Confidence in the primary hypothesis, in [0,1].

  • alternativeHypotheses

    Other OCS hypotheses this same evaluator weighs as possible, each with its own confidence — the intra-claim uncertainty axis.

Shared fields

  • evaluatedAt
  • evaluatedBy
  • rationale
  • evidenceRefs

Place on the record

Origin classifications

A list of competing origin classifications for the same observation.

Code: origin (OriginClaim[])

A list of competing classifications of what the observation was.

Where the verdict comes from

Because records keep every claim without ranking them, the questions people actually ask — which claim to trust, whether evaluators disagree, what the confidence range is — are answered one layer up. Weighing evaluators and producing a single number is Part 5: Scoring.

Part 3 of 5OriginsClassificationPart 5 of 5ScoringEvaluation