Architecture ยท audit record

A spine for accountable AI.

The output is not the record. The record is what lets the output be judged.

Skill Core captures the claim, binds the evidence, preserves the authority boundary, observes the outcome stream, and updates trust when reality answers back.

Pattern language

Eight fields, one audit-readable event.

The architecture is AI-native but not AI-limited. Once a claim has evidence, context, authority, and an outcome trail, it can be scored wherever reality later supplies a verdict.

  1. 01
    Claim

    The model answer, agent action, decision, prediction, or recommendation.

    asserted
  2. 02
    Evidence

    Sources, observations, retrieved context, data, tool calls, and reasoning traces.

    bound
  3. 03
    Authority boundary

    Role, permission, policy, escalation, and the point where review becomes mandatory.

    scope
  4. 04
    Context

    Model, version, source set, date, prompt state, environment, and operating conditions.

    context
  5. 05
    Outcome stream

    The later measurements and events that make the original claim checkable.

    observed
  6. 06
    Reality validation

    The comparison between what the AI claimed and what actually happened.

    resolved
  7. 07
    Trust update

    The change in model, source, or workflow weight after evidence meets reality.

    learned
  8. 08
    Failure semantics

    The explicit reason a claim failed, exceeded scope, degraded, or lost support.

    typed