The DTG Credentials Core Specification
The ToIP working draft that names the objects all of this is about. Not my work โ I read it, build against it, and co-chair the task force that owns the layer it deliberately leaves open.
Decentralized Trust Graph Credentials โ Core Specification, Version 1.0, Working Draft. Decentralized Trust Graph Working Group, Trust over IP Foundation.
Editors: Alberto Leon and Brendan A. Miller (Applied Social Media Lab, Berkman Klein Center at Harvard University), Geoff Turk, Martina Kolpondinos and Drummond Reed (First Person Project). Contributors include Sankarshan Mukhopadhyay, Glenn Gore, and the participants of the DTGWG.
What it defines
Six W3C Verifiable Credential types that create and annotate a trust graph, and four identifier types for the nodes.
Edge credentials, issued in pairs โ one in each direction โ to form a single edge: the VRC (a peer-to-peer relationship) and the VMC (membership in a community).
Annotation credentials, which add assertions to an edge that already exists: the VPC (asserting a persona to a counterparty), the VEC (an endorsement), and the VWC (a witness attesting that an edge is authentic โ for example that both parties proved they were at the same event at the same time).
Invitation credentials, which onboard a new member to a community.
Four identifiers: R-DID for a relationship, M-DID for membership, C-DID for a community, P-DID for a persona. A relationship DID is "not intended to be used for any type of correlation beyond the scope of the two peers" โ intentional correlation is what the other three are for.
The two constructions
This is the part most worth knowing, because it decides what a given deployment may claim.
Pairwise, available to any two parties holding a relationship credential between them, with no community anywhere in the picture. It supports selective disclosure and minimal correlation โ and the spec says plainly that it "does not by itself confer any community-level assurance (e.g., personhood)."
Community-anchored, available when both parties hold membership credentials from the same community. Whatever the community's governance attaches to those memberships โ personhood, when they qualify as personhood credentials โ carries forward into the proof.
An implementation is not deficient for being in the first category. It is in the first category. See OASIS And The Cred Spec.
What it leaves open
"Detailed ZK protocols and registry-ZK interactions are out of scope for this specification." That deferred layer is the charter of The DTG ZKP Task Force.
One standing note it makes that the rest of the industry mostly does not: "Implementations SHOULD make ZKP presentation the default behavior so that users obtain privacy preservation without having to opt in." The default is the opposite of the usual default, deliberately.
Read it at trustoverip.github.io/dtgwg-cred-spec โ and read that rather than this page, which is one person's summary of one vendored copy.