Changes apply to all in-flight assessments. Historical completed assessments keep their original question text via version snapshots.
Every question is tagged at the database level to exactly one construct. This is the chain a consultant or client can show the stakeholder asking “where did this question come from?”
1 · Item
Walk through the path of a single inference request: where it lands, where its inputs and outputs are logged, who can read those logs.
اشرح مسار طلب استدلال واحد: أين يصل، أين تُسجل مدخلاته ومخرجاته، ومن يستطيع قراءة هذه السجلات.
2 · Pillar
3 · Score map
No score map (open-text or unscored item).
No validation evidence captured for this question yet. Click below to ask Claude for an AI-suggested anchor against the curated bibliography in the methodology brief. Suggestions are saved as ai_proposed and require admin review before they appear in any client report.
For multiple choice / yes-no, enter options and score_map as JSON.