DIGITAL ASSET RISK

Crypto products can result in total loss. A rating is not proof of solvency, safety or local availability.

Read the safety standard
VERAQUORUMBuild a comparison

Published methodology

How we will rate self-custody wallets.

The framework starts with key control, transaction security and recovery—not exchange trading metrics.

100

total possible points
subject to evidence gates

Weighting model

Every point has a declared purpose.

Weights are published before testing. They do not change to make a preferred product win.

Key & transaction security30%

Key control, signing clarity, permissions, phishing defenses and incidents.

Code & audit transparency15%

Source availability, audit scope, release provenance and governance.

Recovery & backup15%

Recovery models, backup choices, failure states and user responsibility.

Network & dapp coverage15%

Chains, assets, hardware support and onchain application access.

Privacy10%

Telemetry, data collection, account requirements and third parties.

Usability & support10%

Platforms, accessibility, transaction comprehension and help channels.

Built-in transaction cost5%

Wallet fees, provider markups and separation from network gas.

Expert dossier minimum

Every mature file answers the decision first, then covers scope and legal entity, custody or key control, security, market-specific regulation, total cost, funding and exit, execution or wallet permissions, privacy, support, incident history, alternatives, confidence, change triggers and at least four reader scenarios. Ten distinct analysis modules are the floor, not a substitute for evidence.

Scenario evidence

A scenario is documented with its market, account type, product version, device, asset, amount, time, steps and retained outputs before it is counted as evidence. Public documentation may establish the rule being tested; it cannot stand in for a funded order, withdrawal, recovery or support observation.

Evidence gate

A numerical score is not calculated when critical identity, custody, legal or operational facts cannot be verified. Product convenience cannot average away an unresolved hard risk.

Missing data

Missing capability data is labeled “Not verified.” Missing identity or regulatory evidence can trigger an unrated state. Missing transparency may reduce the transparency component only when the absence itself is decision-relevant.

Community evidence

User reviews can inform experience signals such as support, withdrawals and usability. They never establish regulatory status, security controls or solvency and never modify the expert score directly.

Change control

Material edits carry a verification date and a change note. A new marketing page alone does not count as independent verification. Scores will be re-evaluated when product, entity, fee, security or regulatory evidence changes.

Public and machine-readable parity

The visible server-rendered dossier is the controlling publication. Structured data, RSS, the research index and llms.txt may summarize or navigate that same record, but they cannot introduce a rating, claim or conclusion that is absent from the page readers see.