a2aproject/a2a provides an open protocol that lets opaque AI agents built on diverse frameworks communicate and collaborate without sharing internal memory or tool implementations. Agent communication uses JSON-RPC 2.0 over HTTP(S), enabling secure collaboration across separately hosted applications.
Project overview
It standardizes inter-agent communication so that separately hosted agents can collaborate on long-running tasks without exposing proprietary logic, internal state, or specific tool implementations.
Project type
AI Agent · Infrastructure
Use cases
Automation
Deployment
Refer to project documentation
License
Apache-2.0
Best for
Developers and enterprise teams who need to establish communication and interoperability between opaque agentic applications running on separate servers.
Teams managing AI agents built with diverse frameworks who require a standardized API interaction mode to exchange structured data.
Key capabilities
Defines agent communication using JSON-RPC 2.0 over HTTP(S).
Defines exchange of text, files, and structured JSON data between agents.
Enables agents to collaborate securely on long-running tasks without exposing their internal state, memory, or tools.
Limitations and risks
Self-hosting, local runnability, managed service availability, deployment methods, and data boundary constraints are not documented.
Runtime requirements including GPU, minimum hardware, operating systems, databases, external services, and model requirements are not documented.
Telemetry practices and cost dependencies are not documented.
Getting started
Setup difficulty is unclear because the repository directs users to protocol documentation, a specification, separate SDK repositories, and samples rather than documenting a coherent first-success setup route for the repository itself.
Alternatives and comparisons
Provides a unified multi-provider LLM API, an agent runtime with tool calling, and an interactive coding agent CLI with a terminal UI.
An open-source super agent harness that researches, codes, and creates using sub-agents, memory, sandboxes, and extensible skills.
Provides autonomous AI penetration testing agents that dynamically run code, find vulnerabilities, and validate them with real proofs-of-concept.
GitHub project description: Agent2Agent (A2A) is an open protocol enabling communication and interoperability between opaque agentic applications.
README: The Agent2Agent (A2A) protocol addresses a critical challenge in the AI landscape: enabling gen AI agents, built on diverse frameworks by different companies running on separate s…
README: With A2A, agents can: - Discover each other's capabilities. - Negotiate interaction modalities (text, forms, media). - Securely collaborate on long-running tasks. - Operate withou…
README: - **Standardized Communication:** JSON-RPC 2.0 over HTTP(S). - **Agent Discovery:** Via "Agent Cards" detailing capabilities and connection info. - **Flexible Interaction:** Suppo…