The curated list of Agent2Agent (A2A) protocol projects, templates, tools, and resources — enabling a future of interoperable AI agents.
Agent2Agent (A2A) is an open communication protocol developed by Google to enable secure and standardized interaction between AI agents built by different teams, across different platforms.
It defines a common language and interaction model for agents to:
- Discover each other
- Negotiate capabilities
- Collaborate on tasks
- Communicate securely
Learn more in the Official A2A Documentation
This repo aims to become the central hub for the Agent2Agent ecosystem.
We collect, organize, and showcase:
- Open-source A2A-compatible agent projects
- Ready-to-use Agent2Agent API Templates
- MCP-based A2A Server demos
- Web-based Agents Hub interface
By following the A2A standard, we believe the future of AI will be collaborative, modular, and decentralized.
- Official Resources
- Templates & Demos
- Agents Hub
- Open Source Agents
- Commercial Agents
- Tutorials & Guides
- Research Papers
- Roadmap
- Contribute
| Name | Description | Language | Link |
|---|---|---|---|
| FastAPI A2A Template | Starter project for building A2A-compatible agents using FastAPI. | Python | View Template |
A web-based tool for discovering, testing, and interacting with A2A-compliant agents.
Key Features:
- Agent discovery & management
- Session handling
- Interactive testing
- Pre-configured agent gallery
- Conversation history
Curated list of public A2A-compatible agents.
List of commercial / closed-source agents implementing the A2A protocol.
Coming soon.
Coming soon.
We are actively building the following projects to accelerate Agent2Agent (A2A) protocol adoption:
Reference implementation templates to help developers quickly build Agent2Agent (A2A) protocol-compliant services.
Features:
- Fully aligned with the official A2A protocol specification
- Built-in handling of HTTP request/response formats
- Session management for conversation context
- Structured response and error handling
- Ready-to-use FastAPI implementation with OpenAPI documentation
- Extension points for custom agent logic
We are exploring the integration of Google's Multimodal Conversational Platform (MCP) with the A2A protocol.
This server template wraps A2A-compliant agents as MCP servers to maximize interoperability while maintaining protocol compliance.
Key features:
- Full implementation of A2A HTTP specifications
- Session management for multi-turn conversations
- Structured task handling and artifact generation
- FastAPI-based implementation with OpenAPI documentation
- Customizable agent capability extension points
A web-based tool to help developers discover, test, and interact with A2A-compliant agents.
Key features:
- Agent discovery and management (connect to any A2A-compatible agent via Base URL)
- Interactive API testing with automatic session handling
- Pre-configured gallery of ready-to-use agents
- Conversation history tracking
- Capability exploration within the A2A protocol specification
A growing collection of agents that implement the A2A protocol — including both our own projects and contributions from the open-source community.
The gallery will cover agents from multiple domains:
- Productivity & Work
- Developer Tools
- Healthcare
- Finance
- IoT & Smart Home
- Research
- And more
We welcome contributions from the community!
- Submit your A2A-compatible agent project
- Add tutorials, tools, libraries
- Share research papers or use cases
- Improve templates or demos
Check the Contribution Guidelines to get started.
