Machine-Readable Interface (MRI)

A structured, API-first interaction layer that allows external autonomous agents to discover, evaluate, and transact with a business's services without a human intermediary — built specifically for agent inference, not developer integration.

A Machine-Readable Interface is not an API in the conventional sense. A standard API was built for developer integration: it assumes a human engineer on the other end who reads documentation, writes code to call the endpoint, and handles the response in a custom application. An MRI is built for agent inference: it assumes an autonomous agent on the other end that is evaluating a service as a procurement option, comparing it against alternatives, and making a transaction decision without a human in the loop. The distinction is not the technology. It is the design intent. An MRI exposes pricing, availability, service specification, and transaction initiation in the schemas that leading LLMs and agentic frameworks use to identify and qualify service providers. It is a designed discovery mechanism, not a data access layer. A business with a well-built API but no MRI is accessible to developers. It is invisible to agents. In the A2A economy, that invisibility is a structural commercial liability.

First used: March 2026