test suite scoring framework
Sevnova Research Archive
Version: v1.1
Status: Foundational
This document describes how results from the Home Base AI Test Suite are evaluated and interpreted.
The scoring framework defines how prompts, tests, and evaluation phases are scored, and how outcomes should be reported when systems are tested. It establishes consistent interpretation rules for pass, fail, and diagnostic outcomes while preserving the integrity-first philosophy of the suite.
The framework explains how gating tests differ from diagnostic tests, how phase-level results should be understood, and how evaluation reports may be published without revealing proprietary test prompts.
This document focuses on scoring methodology and interpretation rather than test content.
──────────
──────────