experiment-governance

What problem does it solve?

Maintains and extends the advisory governance layer that classifies experiment usefulness, concentration, noise, and scoped-candidate status from existing paper-research metrics. Use when updating experiment status rules, governance summaries, or operator-facing decision support. Never use for auto-enabling experiments or changing scorer behavior directly.

Core Features & Use Cases

• Transform existing summary metrics into standardized governance statuses and rule triggers for experiments.
• Support operator-facing decision-support widgets and audit trails with passive, advisory conclusions.
• Capture explicit manual notes, defer/reject reasons, and promotion history within a shared audit model.

Quick Start

Review the current experiment metrics and generate a governance verdict with supporting evidence for operator review.