Version 1.0.1

• 21 questions across seven dimensions;
• four evidence anchors per question, scored 0–3;
• dimension totals from 0–9 with four published bands;
• critical cap across environments, verification, and ownership;
• separate Controlled orchestration-readiness requirement; and
• deterministic risks, actions, and 14-day pilot.

Recommended actions now prioritize a lower dimension total before using criticality as a tie-breaker inside the same maturity band. This keeps the next action focused on an observed deficit while preserving escalation when critical and non-critical controls are otherwise equally weak.

Version 1 answers use a dedicated localStorage key and include both storage and instrument versions. Unknown, stale, malformed, or out-of-range payloads are ignored rather than migrated silently.

A change to question meaning, scoring thresholds, critical dimensions, overall calculation, or readiness criteria requires an instrument version change and new acceptance-fixture evidence.

The code and instrument are released under the MIT License. Public contributions should preserve deterministic scoring, privacy boundaries, and acceptance fixtures.