Start Here

Purpose

This file provides the shortest entry path for using the LUMINA-30 incident-review materials.

Use This Repository When

Use this repository when an AI-related incident raises questions such as:

• Could humans still refuse, stop, or override the AI-driven action before irreversible impact?

• Was human refusal operationally effective, not merely nominal?

• Are records sufficient to verify the refusal pathway?

• Did AI-generated output become the sole or primary rationale for final action?

First Path

Start by opening the first-use guide, then follow the three core files listed there.

1. FIRST_3_FILES.html

The three core working files are:

1. LUMINA-30 Boundary Check
2. Incident Quick Assessment Template
3. Boundary Check Result Classification

Before classifying the result, apply the consolidated decision criteria:

• Irreversibility Decision Criteria
  Use this to check who decides, when irreversibility is judged, and what evidence is required.
  

Minimal Review Flow

1. Identify the incident boundary.

2. Check whether irreversible impact was possible or approaching.

3. Determine whether human final refusal was available before that point.

4. Verify whether the refusal pathway was recorded.

5. Apply the consolidated decision criteria.

6. Classify the result without making legal conclusions.

Status

LUMINA-30 is non-binding, descriptive, and intended as a boundary reference for post-incident review.