OAIES Validation-to-Specification Issue Template
Template version: 1.0.0
Published: 2026-07-16
Status: Controlled issue template
Purpose
Convert an independently observed pilot disagreement or limitation into a traceable, testable OAIES specification decision.
Why
Unstructured pilot feedback encourages anecdotal edits and retrospective score optimization. A source-linked issue with an acceptance test and version-impact analysis protects normative stability.
How
Create one issue per distinct root cause after the primary agreement report locks. Do not include participant-identifying or restricted evidence in the public issue.
Issue Control
| Field | Value |
|---|---|
| Issue ID/version | {ISSUE_ID_VERSION} |
| Status | {TRIAGE_ACCEPTED_REJECTED_DEFERRED_IMPLEMENTED_VALIDATED_CLOSED} |
| Source pilot/cohort | {OPAQUE_IDS} |
| Source disagreement/finding IDs | {IDS} |
| OAIES/control versions | {VERSIONS_AND_CONTROL_IDS} |
| Classification | {SPEC_AMBIGUITY_EVIDENCE_AMBIGUITY_RUBRIC_GAP_SCOPE_ERROR_DATA_ERROR_ASSESSOR_ERROR_REASONABLE_JUDGMENT} |
| Severity | {CRITICAL_HIGH_MEDIUM_LOW} |
| Owner/due date | {OWNER_YYYY-MM-DD} |
| Confidentiality | {PUBLIC_REDACTED_RESTRICTED} |
Problem Statement
| Field | Required content |
|---|---|
| Observed ambiguity/failure | {FACTUAL_DESCRIPTION_WITHOUT_REWRITING_ORIGINAL_RATINGS} |
| Competing interpretations | {INTERPRETATION_A_AND_B} |
| De-identified evidence references | {PACKAGE_ARTIFACT_IDS} |
| Affected profiles/impacts | {SCOPE} |
| Safety/compliance/assurance effect | {CONSEQUENCE} |
| Frequency | {COUNT_DENOMINATOR_COHORT_LIMITATION} |
Proposed Resolution
| Field | Value |
|---|---|
| Change class | {NORMATIVE_GUIDANCE_TOOLING_TRAINING_NO_CHANGE} |
| Proposed text/behavior | {EXACT_CHANGE} |
| Why this resolution | {RATIONALE} |
| Alternatives rejected | {ALTERNATIVE_AND_REASON} |
| Backward compatibility | {PATCH_MINOR_MAJOR_AND_CLAIM_EFFECT} |
| Migration/deprecation | {REQUIRED_ACTION_OR_NONE} |
| Cross-control/crosswalk impact | {AFFECTED_IDS_DOCUMENTS} |
Acceptance and Validation
Feature: {SPECIFICATION_BEHAVIOR}
Scenario: {INDEPENDENT_REGRESSION_CASE}
Given {BOUNDARY_AND_EVIDENCE_CONDITIONS}
When {QUALIFIED_ASSESSOR_APPLIES_CRITERIA}
Then {ONE_OBSERVABLE_INTERPRETATION_OR_EXPLICIT_JUDGMENT_BOUNDARY}
| Gate | Evidence |
|---|---|
| Independent technical/editorial review | {REVIEW_IDS} |
| Legal/privacy/security review when affected | {REVIEW_IDS_OR_NOT_APPLICABLE_REASON} |
| Regression fixtures excluding measured pilot as sole confirmation | {FIXTURE_IDS} |
| Public review and dissent | {COMMENTS_DECISIONS} |
| Version/release approval | {DECISION_ID} |
| Future external-cohort validation | {COHORT_ID_RESULT_OR_PENDING} |
Decision
{ACCEPT_REJECT_DEFER} because {EVIDENCE_BASED_RATIONALE}.
State explicitly whether external validation remains pending. Closing an editorial issue is not proof that real independent assessors now agree.
Tradeoffs
| Decision | Benefit | Cost |
|---|---|---|
| One root cause per issue | Clear ownership and regression | More issue records |
| Exact proposed text | Reviewable semantic impact | Requires specification-author effort |
| Future-cohort validation | Tests generalization | Delays evidence of resolution |
Anti-patterns
- Editing normative text directly from one disputed case.
- Marking an issue validated because internal authors agree.
- Hiding rejected interpretations or dissent.
- Using the measured item as the only regression test.
- Closing a normative issue without version and claim-impact analysis.
Enterprise Considerations
Public issue trackers must not contain restricted workpapers, vulnerability details, personal data, or participant identity. Preserve a controlled trace from redacted issue to source evidence for authorized reviewers.
Checklist
- Primary metrics locked before issue creation.
- Root cause and affected controls are explicit.
- Evidence references are de-identified but traceable.
- Proposed resolution and rejected alternatives are recorded.
- Acceptance test is observable.
- Version, migration, and active-claim effects are approved.
- Future external validation is complete or clearly pending.
Authoritative References
- OAIES Blind Inter-Assessor Methodology.
- OAIES Lifecycle and Assessor Methodology.
- Semantic Versioning 2.0.0.
Changelog
| Version | Date | Change |
|---|---|---|
| 1.0.0 | 2026-07-16 | Established traceable pilot-to-specification issue and validation fields. |