Research and Standards
OAN is not only a website or a set of services. It is a research-backed infrastructure program with papers, public specifications, slide material, standards-facing drafts, SDKs, and running implementation. This section collects the deeper material for readers who want to evaluate the model beyond the introductory Docs.
Most users do not need to start here. But researchers, protocol reviewers, standards collaborators, and technical decision makers should find the primary references here.
The purpose of this section is to make OAN reviewable. A reader should be able to move from the public story, to the technical architecture, to semantic Discovery research, to standards-facing drafts, and then back to the running implementation. That loop is what distinguishes a serious infrastructure effort from a one-off demo.
How To Read the Materials
| Material | Best for | Main question |
|---|---|---|
| White paper | Vision, ecosystem positioning, adoption. | Why does OAN exist? |
| Yellow paper | Protocol architecture and trust model. | How does OAN work technically? |
| GRAIL paper | Semantic discovery and retrieval quality. | How can Discovery find better matches? |
| IETF drafts | Standards-facing interoperability discussion. | How can the model be expressed for broader review? |
| Slides | Fast narrative, diagrams, demos, implementation evidence. | How do I explain OAN to others? |
The materials are complementary. The white paper motivates the problem. The yellow paper defines the trust architecture. GRAIL explores semantic Discovery. IETF drafts translate parts of the model into standards-oriented language. Slides connect concepts to diagrams and implementation status.
Do not expect the materials to share the same chapter structure. They serve different audiences. The important alignment is conceptual: trusted resource identity, governed node authority, Root-verifiable package publication, authorized Discovery, semantic retrieval over verified material, and verification before native protocol use.
Why This Matters
Agent-resource infrastructure sits between several worlds: identity, governance, discovery, semantic retrieval, agent protocols, APIs, package distribution, and security verification. A credible system needs more than a product page. It needs public argument, technical specification, implementation evidence, and a path for external critique.
OAN’s research and standards material is meant to support that critique.
It should also help external projects decide where OAN fits. OAN is not trying to become every agent interaction protocol. It is a trust and Discovery layer that can sit before MCP, A2A-style task exchange, APIs, skills, and domain-specific protocols.
Primary Links
- OAN white paper
- OAN yellow paper
- GRAIL semantic discovery paper
- AONA paper
- DarwinNet paper
- IETF draft: OAN resource identity and discovery
- IETF draft: Agentic Overlay Network architecture
- IETF draft: Efficient Agent Discovery profile
- did:oan DID Method Specification
Research Themes
The recurring themes are:
- trusted resource identity;
- governed infrastructure authority;
- Root-verifiable package publication;
- authorized-domain scope;
- Discovery over verified resources;
- semantic retrieval without trust bypass;
- verification before invocation;
- protocol neutrality across MCP, APIs, skills, and agent services;
- staged adoption and third-party node participation.
These themes also map to the implementation. The papers should not be read as isolated theory; they connect to repositories, SDKs, services, website flows, and trial-network material.
This connection gives reviewers concrete things to test. If a paper says Discovery is authorization-aware, the implementation should show domain-filtered behavior. If the architecture says clients verify before use, the SDK should expose verification helpers. If the slides claim benchmark progress, the demos and test scripts should make the claim inspectable.
Further Study
Start with the white paper if you want the story. Start with the yellow paper if you want the protocol architecture. Start with GRAIL if you care about semantic Discovery. Start with IETF drafts if you care about interoperability language. Use the slides when you need diagrams for discussion or presentations.
For a complete review, read in two passes. First, read the white paper and slides to understand the narrative. Then read the yellow paper and drafts while checking the website, SDK, and service behavior against the trust claims.