U
Criterion 1
Purpose and utility
What the project does, what the token is used for, and whether that use provides a recognisable lawful benefit.
Loading…
Transparent Shariah research
Final authority
FARUQ prepares a traceable evidence record. Qualified scholars review that record, apply the approved methodology and make the final Shariah determination. The engine does not issue a fatwa and cannot replace scholarly judgment.
Engine
Gathers and checks facts
Scholar
Reviews evidence and interpretation
Decision
Published with scope and attribution
Eight research areas
Criterion 1
What the project does, what the token is used for, and whether that use provides a recognisable lawful benefit.
Criterion 2
The project’s actual activities and material exposure to prohibited products, services or revenue.
Criterion 3
Whether lending, borrowing, returns or other contractual mechanics create an interest-based obligation or entitlement.
Criterion 4
Whether the product contains an evidenced wagering structure—not merely volatility, trading or speculation.
Criterion 5
What the holder owns or controls, which rights attach to the token, and how possession is established.
Criterion 6
How the asset is delivered or settled and whether material uncertainty remains in that process.
Criterion 7
Where protocol revenue and holder yield come from, and the contractual basis for any return.
Criterion 8
Whether the token represents or is redeemable for another asset and whether specialist review is required.
Missing information never becomes a pass. Price, popularity, token name, privacy features, volatility and consensus mechanism do not determine a decision by themselves.
Evidence gathering
Every coin walks the same seven steps. Code gathers and checks the facts; AI reads documents and quotes what they say; qualified scholars review the complete record and give the final decision. The engine supports that judgment—it does not make it.
every coin walks these seven steps
Identify the coin
the exact token, not the ticker
Collect the sources
attributed project documents
Read the documents
two blind OpenAI passes
Cross-check every fact
source, meaning, identity and freshness
Organise facts under 8 criteria
versioned questions, no score
Check evidence dependencies
mandatory facts and unresolved gaps
Scholar review and decision
qualified judgment on the complete record
Open a step to see exactly what it does.
The case is bound to one provider ID, network and contract address. Tokens that share a ticker remain separate subjects, and evidence collected for one cannot be used for another.
Deterministic checks
The collector retrieves project documentation and stores each source with its URL, retrieval date and a hash of the exact text read. An unreadable source is recorded as a failure, never as clean evidence.
Deterministic checks
Two OpenAI passes read the same stored passages separately. They extract factual claims and the exact passages supporting them. Their answer format contains no field for a religious verdict or permissibility judgment.
AI reading
Code checks source identity, freshness, quote presence and whether the quote supports the extracted claim. Ambiguous or conflicting material is flagged for human review instead of being silently resolved.
Deterministic checks
Validated facts are organised under the current versioned criteria. Missing information remains missing, criteria are not averaged into a compliance score, and this stage does not issue a religious ruling.
Deterministic checks
The engine checks whether the required facts and supporting sources are present for the requested scope. It records gaps, conflicts and scope limits so the scholar can see exactly what is established and what is not.
Deterministic checks
A qualified scholar reviews the sourced facts, unresolved questions and applicable methodology before giving the final decision. Evidence expires after 90 days; stale or disputed cases are held for renewed review rather than presented as current conclusions.
Qualified scholars
Scholar review
The scholar finds the asset compliant within the exact scope reviewed.
The conclusion depends on clearly stated limits or conditions.
The scholar identifies a decisive conflict with the applied methodology.
Material scholarly or factual uncertainty prevents a clear conclusion.
The evidence is incomplete, stale, conflicting or outside the approved scope.
The reviewer and review date are shown. FARUQ does not create anonymous endorsements.
Holding an asset does not automatically cover staking, lending, margin, derivatives or a particular venue.
The published decision points back to the facts, sources, methodology version and unresolved limitations.
Project documents are stored with their URL, retrieval date and a hash of the exact text read.
Every extracted fact must retain the passage that supports it. A matching AI answer is not proof by itself.
Identity, source quality, freshness, wording and factual conflicts are checked separately.
Unreadable, missing, ambiguous or conflicting evidence is shown as a gap and routed for human review.
Evidence expires after 90 days and published records retain the methodology version applied. A failed refresh does not silently reuse old evidence as though it were current.
Earlier assessment versions remain available when facts, methodology or a scholar’s conclusion changes. Corrections and disputes do not erase the history.
Important: two blind OpenAI readings are a consistency check over the same documents—not two independent sources and not a scholarly opinion.
Current coverage
Crypto assets at asset level. Each result must name the exact token, network or contract identity and the activity scope assessed.
Stocks, ETFs, forex, margin, derivatives and product-specific staking or lending rules. Equity financial ratios are not transferred to crypto by default.
No mixed-revenue tolerance, grading score or unsupported exception is applied unless it exists in an approved, versioned methodology.
FARUQ records research and attributed scholar decisions. It does not claim accreditation, manufacture endorsements or present automation as a fatwa.
The current bundled evidence framework is 0.2.0-draft. Detailed release records and source limitations are retained here for auditability without interrupting the public explanation above.
Every available asset record shows its sources, factual gaps, scope and review state.