The research pillar for AxoneOS — architectural, comparative, and institutional work on what governs agents when transports come and go. Begin with the MCP-compatible architecture pillar; future pillars extend the same thesis to new domains.
Every protocol that wins a year loses to a better one the next year. What does not lose is the registry of who is allowed to do what, under which rules, with which settlement, and how disputes are resolved when those rules break. This pillar treats the architecture, the institution, and the auditable trail that outlives any single wire as the durable substrate of multi-agent systems — see also the foundation thesis and the MCP-compatible architecture pillar.
Today's first published pillar argues that the Model Context Protocol standardizes the wire format but not the institution. The AxoneOS research entry lays out why a governance layer — not another transport — is the part that must outlive MCP.
Concise references complement the research pillars with practical perspectives on institutional data sovereignty, integration, and privacy.
Why record-retention obligations and audit-grade access controls outlive any single model vendor or cloud — and how Axone Zones ground that sovereignty on chain.
Read the pillar →A concise guide to external adapters, uncorrelated identities, credentials, and off-chain effects, with each pattern’s privacy implications.
Read the reference →Catch the MCP-compatible architecture pillar, or revisit the long-form articles that motivate it — the blog is the public research surface for governance and architecture on AxoneOS.