Ovragrid Systems: recorded account_prioritization decision

The recorded sections show stored decision data. The replay panel can compare that preserved context under selected logic without changing the record.

Decision class

account_prioritization

Output

DO_NOT_PRIORITIZE

Score and threshold

score 35 / threshold 75

Decision boundary T(d)

2026-08-07T13:58:16.000000Z

Occurred at

2026-08-07T13:58:16.000000Z

Recorded at

2026-08-07T13:58:23.000000Z

Event source

relaybridge-scoring

Workflow version

v4.2

Decision event id

ds-0167-decision-1

Confidence: not provided. No confidence value was recorded for this decision, so no number and no bar is shown for it.

Historical logic identity

Logic version (label)

v5.1

Artifact id

logic-account-prioritization-v5.1

Artifact hash (identity)

b5a6f33dd592fe804d9b59e081639c6527a9c4ca2727cb6f850c6888b3ea7a3b

Artifact schema version

1

Evaluator version

evaluator-v1

The version label is metadata. The artifact hash together with the evaluator version is the identity that makes this decision exactly replayable.

Declarative ruleset as recorded

Threshold

75

Output at or above threshold

PRIORITIZE

Output below threshold

DO_NOT_PRIORITIZE

Missing-value behavior

no_match_contributes_zero

Activated at

2026-06-01T09:00:00.000000Z

Deactivated at

none

Activation status

current

Positive weights total 72, below threshold 75; the score cannot reach this threshold under evaluator-v1. A successful evaluation under this artifact can therefore only output DO_NOT_PRIORITIZE. This is a property of the recorded artifact, not of this account, and the bound itself is not a failure. It says nothing about whether this decision reconstructs; any reconstruction outcome is reported in its own section.

Preserved historical context

Each state is a recorded fact: consumed means the historical logic referenced the input, available but ignored means it was present and no factor referenced it, unavailable means it was explicitly recorded as unavailable at the boundary, and absent means the historical logic references a key the preserved context has no entry for. A consumed input whose rule did not match contributes 0; every other state reads not consumed rather than 0.

Explanation (recorded narrative, not evidence or logic)

No explanation recorded.

The explanation presents the recorded basis for this decision. It is not the authoritative record: the evidence, the input states, the ruleset, and the score above are. It is stored exactly as it was submitted and is never regenerated or rewritten here.

Downstream action

No action recorded for this decision.

Recorded account outcomes

These are outcomes recorded for this account. The attribution column shows, for each recorded outcome version, whether a policy-based link to a decision has been evaluated and what that evaluation recorded. Details for each outcome version follow the table in the same order.

A window shown only as a length comes from a record that states the window's length and nothing about whether the window is open or closed, and this page does not infer it; for such a record, and for a closed window, an observation of no is a recorded negative observation. In an open window, no means nothing was recorded as of the stated time, not a negative result. Unknown means the observation itself was not recorded. An attribution is a policy-based link to a decision, not proof of what caused the outcome. An attribution history lists earlier results as recorded, and only the effective one is the outcome version's current attribution.

Replay: the preserved historical context under selected current logic

Replay produces a counterfactual. This decision did not occur. Replay evaluates the historical context preserved at the decision boundary under a selected current logic artifact. No present-day evidence enters it, and a replay result is never stored: no table, column, event, or cache holds one.