A source-grounded research layer that explains methodology, standards, product boundaries, and re-screening reasons without issuing personal religious rulings.
The roadmap is explicit: AI should summarize, classify, route, and explain, while hard rules handle risk limits, jurisdiction gates, sanctions checks, and permissions. That keeps the assistant useful without turning it into a black-box authority.
Every answer should point back to published methodology, AAOIFI-aligned screening references, OIC Fiqh Academy principles, product boundaries, and the relevant asset or protocol evidence.
The assistant can explain why a token passed, failed, or needs review under the operating framework. It must not provide a personal fatwa, individual investment advice, or country-specific legal advice.
Ask
The assistant retrieves methodology gates, current classification, and source notes before drafting an explanation.
Ground
Responses should link to methodology, governance, and asset evidence instead of inventing religious authority.
Escalate
Unknown assets, jurisdiction-sensitive questions, and personal religious rulings should route to human or scholar review.
