Platform disclosures
Make the operator, fees, access restrictions, conflicts and support routes easier to understand.
Discuss this reviewFor prediction-market businesses
Clearer practices.
Evidence people can inspect.
A transparency review helps you find gaps in a defined part of your product, explain what you do and document the improvements. We agree the question and method before testing.
Requests are free. We confirm availability, scope, price and terms before you order work. Reviews provide factual and technical findings; they do not grant regulatory permission or a general safety guarantee.
Choose the question that matters
Contract rules and resolution
Find ambiguous outcomes, unclear sources, time cutoffs and gaps in the dispute process.
Discuss this reviewMarket data and analytics
Check a defined sample for timestamps, missing observations, calculation reproducibility and clear limitations.
Discuss this reviewWhat you receive
- A fixed scope, sample and dated evidence report.
- Findings against the agreed criteria, including gaps and uncertainty.
- A practical improvement plan and a findings walkthrough.
- A factual correction window and a separately scoped retest if useful.
Scope and price
For planning, a single-module review is expected to cost €1,500–€3,000 excluding VAT. This is an indicative range, not an automatic quote.
The final price depends on the product, sample, access and reviewer work. New data licences, specialist legal advice and retests are quoted separately. Fees pay for the work and report, regardless of the findings.
Published criteria
Proposed rubric 0.1.0 contains 24 checks. Each engagement pins the version, sample and acceptance thresholds before testing. These are Vultax criteria, with no claim of regulatory approval or accreditation.
Download the complete criteria as JSONPlatform disclosures: eight checks
D01. Identify the operator and product
Compare claimed legal operator, registration identifier and product role against an authoritative register or equivalent primary record. Record jurisdiction and exact product.
Acceptance: Identity and role are corroborated; brand and legal entity are not conflated.
D02. State regulatory claims precisely
Record exact licensing and registration claims. Check relevant official records for the named entity, permission, product and jurisdiction.
Acceptance: Every material status claim is supported with scope and source date; no global legality implication.
D03. Explain access restrictions
Compare public eligibility, territorial and age restrictions across onboarding and terms. Describe whether access enforcement was observed.
Acceptance: The sample contains consistent disclosures; untested enforcement remains explicitly unassessed.
D04. Explain charges and incentives
Reconcile fees, payout deductions, liquidity incentives and material paid relationships across the chosen interface, API and terms.
Acceptance: Selected examples reproduce the disclosed fee arithmetic and identify exclusions.
D05. Explain funds and settlement roles
Record who holds funds, how settlement is described, withdrawal conditions and any upstream venue relationship. Inspect supporting documents without requesting keys.
Acceptance: Material public claims are supported or qualified, and unsupported insurance or guarantee claims are absent.
D06. Provide a usable complaint and correction route
Inspect published complaints, dispute escalation and factual correction processes, timing and responsible role. No unsolicited test messages.
Acceptance: Routes and responsibilities are stated and usable from the inspected interface; response performance is claimed only if tested.
D07. Disclose the evidence behind assurance claims
Check cited third-party audits, independence statements, publication dates and scope limitations against original reports.
Acceptance: Assurance claims match the exact report, version and scope; no badge substitutes for a licence.
D08. Make changes and conflicts visible
Inspect change logs, effective dates, conflict disclosures, material terms changes and notice commitments.
Acceptance: Material terms have identifiable versions and known conflicts are explained without implying that disclosure eliminates the conflict.
Contract rules and resolution: eight checks
R01. Define each sampled outcome
Freeze sampled contract text, identifiers and answer options before evaluation. Test concrete boundary examples.
Acceptance: An independent reader can determine which stated outcome applies to each tested example.
R02. Name an authoritative resolution source
Record the source, backup hierarchy and what happens if it changes, disappears or corrects its data.
Acceptance: The source and fallback order are explicit and the sampled source is identifiable.
R03. Specify time and data cutoffs
Record timezone, event deadline, source publication cutoff and late or revised data treatment.
Acceptance: All sampled boundary cases have an explicit rule or a disclosed unresolved limitation.
R04. Explain cancellation and exceptional outcomes
Test postponement, abandonment, contradictory sources and other relevant exceptional cases.
Acceptance: Relevant exceptional outcomes and payout treatment are stated; absence is recorded as a gap.
R05. Explain resolution and dispute authority
Identify who proposes, decides and can challenge a resolution, deadlines, costs and conflicts.
Acceptance: A participant can identify the decision route and challenge conditions from the frozen rules.
R06. Preserve rule versions
Compare visible rules, linked terms, API metadata and retained versions for the chosen sample.
Acceptance: Material conflicts are absent or resolved, and amendments are traceable with effective dates.
R07. Reconcile a settlement sample
Select all eligible settlements in a preregistered bounded population or a reproducible sample; compare outcome, source and rule version.
Acceptance: Sampled outcomes follow the recorded rule; unsettled or unobservable cases are retained as limitations.
R08. Review ambiguity independently
Use two readers for the same fixed edge cases; record disagreements and resolution before any mark decision.
Acceptance: Material reader disagreement is resolved by a documented rule or remains a blocking gap.
Market data and analytics: eight checks
M01. Identify the measured data
Record venue, contract, outcome, unit, timestamp semantics and source/collection clocks for each sampled endpoint.
Acceptance: Every displayed number can be tied to the correct contract and measurement definition.
M02. Measure freshness
Measure age distributions over an agreed window with clock checks and separate capture, API and screen latency.
Acceptance: Predeclared thresholds and percentiles are reported; a median is not presented as a maximum.
M03. Report gaps and outages
Measure missingness, sequence gaps where available and failures at the storage/output layer. Preserve exclusions.
Acceptance: Gaps are counted; unknown values stay unknown; carried-forward values are labelled.
M04. Explain prices and tradeable size
Reconcile bid, ask, last trade, midpoint, depth and hypothetical execution cost for a fixed sample and size.
Acceptance: Displayed labels and examples match what was measured; an illustrative quote is not claimed as an executed fill.
M05. Reproduce fees and computed metrics
Recompute selected fee, P&L, return or score examples using stated inputs, exclusions and version.
Acceptance: All in-scope material arithmetic reproduces within the preregistered tolerance; unsupported history stays unverified.
M06. Define populations and selection
State which wallets, contracts, markets and times are included, the sample selection and known coverage bias.
Acceptance: The reported result is limited to the measured population; a wallet is not treated as an identified person.
M07. Preserve reproducible evidence lawfully
Check calculation version, source artefact hashes, reviewer access and permitted retention and publication.
Acceptance: Another authorised reviewer can reproduce the in-scope result; a hash alone is not represented as access to the data.
M08. Present uncertainty without overclaiming
Check whether risk flags, backtests, simulated fills and heuristic scores have accurate labels and limitations.
Acceptance: No sampled statement turns a statistical flag into proven misconduct, a simulated result into a return, or a score into legal compliance.
An improvement service
We work through findings with the participant and preserve the evidence behind corrections and retests. Factual corrections are free. Directory inclusion and editorial coverage do not depend on buying a review.
Before an engagement, we agree confidentiality and any public summary. A confidential report does not create a public verification mark.
Public marks are not yet available
The programme is in preparation. Public marks require actual independent reviews, appointed decision makers, an appeal route and a current status record. Requesting or paying for a review cannot produce a badge automatically.
We disclose conflicts. Our own products, related parties and commercial competitors cannot receive an independent mark through this initial programme.
Vultax is operated by OmniOS OÜ. A scoped review is distinct from a legal opinion, a statutory audit, accredited gambling-software testing or a regulatory examination. Specialist work needs its own competent provider and agreement.
Tell us what you would like to improve