Attestation
Scope
A signed statement that somebody checked something about a claim and this is what they saw. Attestations are how level 2 exists at all, given that most registers are a web page and not an interface.
The document
{
"rvw": 4,
"claim": "fba0fc855026ba9758a1bbb2a41b3a4a5ebe54be5aca29b07c034d5c0e04ae10",
"attester": "0x9d08da60aab239e1f6acd6d2617aa37e5c4e13ca",
"method": "register-page",
"result": "match",
"evidence": "sha256:5b19b99093d3d45892eb2351c909627fe8db0c144009edc9048d952c757fb9e8",
"source": "https://register.example.qld.gov.au/licence/4102938",
"checked": "2026-11-09T09:14:00Z",
"expires": "2026-12-09T09:14:00Z"
}Signed by the attester as EIP-712 typed data. evidence is the digest of whatever was examined, such as the archived register page, so that a disputed attestation can be argued about against a fixed artefact.
Methods
| Method | What the attester did | Automatable |
|---|---|---|
| register-page | Fetched the public register page named in the claim and compared licence, status and legal name. | Yes, and it should be |
| authority-file | Found the licence in a file published by the authority under RVW-5. | Yes |
| domain-binding | Resolved the RVW-2 record and matched the key. | Yes |
| document | Was shown a licence document directly by the subject. | No |
| in-person | Met the person and matched them to the credential. | No |
Results are match, mismatch or not-found. A mismatch is published like any other result: an attestation layer that only records good news is worthless.
Weight
An attestation is worth what its signer is worth to the party reading it, and no arithmetic here changes that. Three attestations from one operator are one opinion. One attestation from a professional body is not the same object as one from an anonymous key, and the protocol does not pretend otherwise by averaging them.
Consumers are therefore expected to carry their own list of attesters they care about, and to show who attested rather than how many did.
Revow operates no trust list
This is rule four and it is the part most likely to be argued with, because a blessed list would make the layer immediately useful. It would also make Revow the gatekeeper, which is the thing this lab exists to avoid, and would turn every future dispute into a lobbying exercise against one list.
The honest consequence is that level 2 is weak until either a consumer has picked its attesters or an authority removes the need for the layer. That is a real cost and it is accepted deliberately.
Limits
A register page can change between the check and the reading, which is why attestations expire and why their expiry is short. Automated checks are only as good as the page they parse, and a page that is rendered by script may not be fetchable at all. Who attests at scale without becoming a platform is open problem OP-6.