Your learning identity should outlive any single platform. The protocol defines portable identity, verifiable credentials, and guaranteed export — the trust layer stays open; the product layer competes.
A Learning Identity starts simple but is designed to progress toward full user ownership. Your learning history, skill graph overlay, evidence vault, cognitive profile, goals, memory, and sharing permissions belong to you — retention should come from being genuinely useful, never from data lock-in.
Credentials combine issuer digital signatures, evidence hashes for private material, revocation status, versioning, audit logs, and selective disclosure where supported. They interoperate with current standards where appropriate — cryptographic proof replaces, rather than decorates, trustworthy assessment.
Export your record in portable schemas, grant scoped third-party access with expiry and revocation, and request deletion where legally and technically applicable — all self-service.
The protocol evolves in the open: public RFCs, semantic versioning, independent review, and academic, industry, and community participation — with a foundation as the long-term home. Core trust verification is never made proprietary to create lock-in; the commercial moat lives in product experience, hosting, compute, models, and integrations.
Current engineering references in this repository: