Decision Evidence

Elora Decisions

A review layer that turns related operational events into a coherent account of what happened, why it mattered, which authority applied, and how the outcome was reached.

One Decision, Many Events

A decision may span observations, warnings, policy checks, interventions, retries, and a final outcome. Elora Decisions preserves those events as one bounded lifecycle rather than presenting an unstructured event stream.

Decision Record

Records connect actor context, originating surface, evidence references, policy state, concerns, authority checks, and the final disposition.

Decision Timeline

Parent decisions can retain ordered child events, including research stages and runtime interventions, while preserving their individual provenance.

Across the Engine

Decision Domains

Research, runtime, memory, learning, security, worker, and resource-governance activity can be presented through a common evidence model without pretending they share the same authority.

Connected Evidence

Decision histories may reference replay records, Observer reports, threat Cases, runtime telemetry, learning evidence, and the policies applicable at the time.

Learning State

Interfaces distinguish what Elora is learning now, what has been learned with supporting evidence, and what remains an active research objective.

Consistent Presentation

The canonical record can support pages, status messages, exports, and future notifications without creating conflicting versions of the same event.

Authority Boundary

Elora Decisions explains and links decisions; it does not authorize them. Replay remains the evidence trail, and consequence-bearing actions still require valid identity, policy, admissibility, and commit-boundary authority.