# About Agent2Agent Protocol

## The shift to agentic commerce

The way people shop is moving into AI. Discovery, comparison, checkout, and even
post-purchase monitoring increasingly happen inside AI assistants rather than on
a search results page or a brand's own site. This isn't a fringe behavior:
[73% of consumers are already using AI in their shopping journey](https://commercetools.com/blog/ai-trends-shaping-agentic-commerce).
And the money is following — Accenture estimates that by 2030,
[more than 30% of online commerce will run through AI agents, close to $3.1 trillion in transactions](https://www.accenture.com/au-en/blogs/banking/agentic-commerce-payments).

### From scraping to brand-controlled protocols

Until now, AI assistants learned about products by scraping the open web –
guessing at prices, availability, and specs from whatever pages they could read.

That era is ending.

Brands and the platforms behind these assistants are
converging on **agentic-commerce protocols**: structured, brand-published feeds
and endpoints that tell an AI exactly what you sell, what it costs, whether it's
in stock, and how to complete a purchase. The difference is control – the brand,
not a scraper, becomes the source of truth.

The catch is pace. Several of these protocols have emerged in just the last year,
each backed by different platforms, and they ship new releases frequently. Staying
compliant is an ongoing commitment, not a one-time integration. And each protocol
takes a different approach.

## What is A2A?

A2A – the Agent2Agent Protocol – is an open standard that lets AI agents built on different frameworks, by different vendors, discover each other and work together. Where a protocol like MCP connects an agent to tools and data, A2A connects an agent to other agents. An agent publishes an Agent Card at a well-known address describing who it is, which skills it offers, how to reach it, and how callers should authenticate. Another agent reads the card, picks a skill, and hands over a task.

It solves the same problem every shared standard solves. Without one, every pairing of agents needs its own custom integration, and the number of integrations grows with every new agent. With A2A, an agent is described once and can be called by any compliant client.

## How it works

The Agent Card is the front door. It lists the agent's skills, the interfaces it serves, and the security schemes it expects. A2A defines three bindings for those interfaces – JSON-RPC, gRPC, and HTTP+JSON/REST – and a card can declare more than one. Work is organized into tasks: a client sends a message, and the agent tracks the result as a task that moves through a defined lifecycle, so long-running work is a first-class idea rather than a workaround.

Versioning is explicit. Clients state the version they speak with an A2A-Version header, and servers that cannot honor it are expected to say so with a defined error. The current published specification is A2A 1.0 (we validate against the 1.0.1 release), and the specification still provides for 0.3 clients, so the check supports both.

## Who's behind it

Google originated A2A in April 2025 and contributed it to the Linux Foundation, which announced the Agent2Agent project in June 2025 under vendor-neutral governance. The announcement carried statements of support from Google Cloud, AWS, Cisco, Microsoft, Salesforce, SAP, and ServiceNow. The project is Apache 2.0 licensed and developed in the open.

## Where it fits

A2A is a connectivity layer, not a commerce protocol. It is how agents find and call one another; commerce protocols such as UCP and AP2 can ride on top of it, declared as extensions in the Agent Card. For a brand, that makes the card a useful place to be accurate: if an agent acting for a shopper discovers yours through A2A, the card is the first thing it reads.

You can [run a completely free check](/) to see whether your agent's card and live interfaces measure up against the published A2A spec right now. And if you would rather not hand-build and maintain that surface as the specification evolves, [CorgiMaps](https://www.corgimaps.com) connects to your existing data and exposes protocol-compliant endpoints for you, with no changes to your fulfillment stack.