The Decision Log. The audit spine of deployment.

Layer 01 — What gets recorded

The full decision snapshot.

Each log entry preserves the complete context of one cycle's deployment decision:

• Gate / DD State — the capital-state row used; verifies capital-state routing.
• Final Tier # / Name — the authorized throttle tier; the core deployment authority record.
• Authorized Cycle Pool % — the pool before exposure adjustment; baseline for the exposure math.
• Active Open Risk % — downside still exposed; the primary smart-capacity burden.
• Smart Suggested vs Effective — recommendation versus final decision; makes override use visible.
• Total Active Risk After New — open plus fresh deployment; the final exposure reality check.
• Override justification — required for accountability whenever the exception layer is used.

Layer 02 — Why it exists

Memory flatters. Logs don't.

Weeks later, the question is never 'did it work?' — it's 'was the decision coherent given what was known?'. The log is what lets overrides, gate routing, and exposure decisions be audited as a pattern. Skipping the log is a named edge case: without the snapshot, you cannot later audit whether decisions were coherent.

Layer 03 — The log's second life

Deployment history becomes the system's memory of itself.

Beyond auditing individual directives, the accumulated log is a dataset: directive sizes over time, override clusters, mode usage across regimes. Quarterly reads of the log answer questions no single cycle can — whether deployment discipline is drifting, tightening, or holding — making the audit spine double as a longitudinal instrument.