Skip to content

Classify the residual gap against finalized UTC usage #16

Description

@iam-brain

Question

Compare legacy and shadow lineage totals against finalized OpenAI UTC daily usage from July 9 through July 12, 2026, plus ordinary non-fork days. Classify residuals as unavailable history, unsupported event shapes, UTC/local attribution differences, accounting semantics, containment, or ledger defects, and make evidence-backed corrections without tuning to a fixed error percentage.

Acceptance

Require material aggregate improvement over legacy, no material regression on previously accurate ordinary days, and an evidence-backed classification for every large residual. Treat an OpenAI day as provisional until finalized. Include the known interleaved Ultra/normal session, active and archived history, and referenced ancestors outside the normal lookback. Never publish private rollout contents.

Metadata

Metadata

Assignees

No one assigned

    Labels

    wayfinder:researchAFK research ticket on a Wayfinder map

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions