A canonical reference identity framework for software and AI agents.

OSAI is an upstream, model-agnostic framework for naming, distinguishing, and referencing software and AI agents before downstream systems apply authority, attestation, audit, policy, or enforcement. It is not IAM, not authentication, not authorization, not a runtime control layer, not certification, and not a governance platform.

What this is

  • An upstream identity and reference framework for software and AI agents.
  • A controlled publication series intended for citation, critique, comparison, and institutional discussion.
  • A document-first sequence moving from framing, to bridge concept, to companion technical artifacts.

What this is not

  • Not a product, service, platform, credential system, or agent marketplace.
  • Not an access-control, orchestration, runtime-governance, or certification system.
  • Not a claim of authority, ownership, operational control, or implementation maturity.
Reading path

How to read the publication series

1

Framing artifacts

Start with the Abstract & Preface, Reader’s Orientation, Document Control, and Standards-Facing Memo.

2

Bridge package

Read the Technical Concept Note and companion artifacts to see the separation of identity, authority, attestation, resolver posture, and narrow demonstration.

3

Boundary posture

Use the Namespace and Vocabulary Posture Note to interpret namespace, vocabulary, issuer scope, collision handling, and registry non-operation across the Layer 2 document family.

4

White paper

The Layer 2 White Paper is the issued public conceptual reference document for identity and naming as infrastructure for artificial intelligence systems.