A2A Protocol - Agent智能体通信开放标准 A2A Protocol - Agent智能体通信开放标准 A2A Protocol - Agent智能体通信开放标准 Home Documentation 规范 Resources 博客 关于
CTRL K
    CTRL K
      • A2A Protocol Technical Documentation
        • Protocol Introduction
        • Core Concepts
          • AgentCard
          • Task
          • Artifact
          • Message
          • Part
        • Quick Start
        • Getting Started
          • Environment Setup
          • Running Examples
          • First Agent
        • Code Examples
          • JavaScript Examples
          • Python Examples
        • Demo Application
          • UI Component Library
        • A2A Protocol Frequently Asked Questions (FAQ)
      • Specifications
        • Core Protocol Specification
        • Agent Discovery Mechanisms
        • Relationship & Integration with MCP
      • A2A Protocol Blog
        • A2A and MCP: Powerfully Combined Agent Protocols
        • A2A Protocol 2025 Development Progress
        • A2A Protocol Enterprise Practice Case Study
        • A2A Protocol 2025.1 Version Released
      • About Agent2Agent.Info - The Agent2Agent Protocol Technology Hub
      • 404 - Page Not Found
      • More
      • 文档 ↗
      • 技术规范 ↗
      • 关于
      • English
      • 中文
      Open · Secure · Efficient

      Enabling Agent Interoperability in the AI Era 
      The Open Standard for Agent Communication & Collaboration

      An open protocol from Google & 50+ partners, enabling agents across frameworks 
      to communicate, interoperate, and collaborate securely.

      Get Started Specifications

      Agent Discovery

      Discover and identify agent capabilities via the Agent Card mechanism for automated service discovery and collaboration.

      Agent Discovery

      Standardized Interfaces

      Standardized interfaces (JSON-RPC, HTTP, SSE) supporting synchronous, asynchronous, and streaming communication.

      Standardized Interfaces

      Task Management

      Complete task lifecycle management, supporting long-running tasks, real-time status updates, and result synchronization.

      Task Management

      Enterprise-Grade Security

      Supports robust identity authentication (e.g., leveraging standards like W3C DID), end-to-end encrypted communication, and enterprise-level authorization mechanisms.

      Multi-Modal Support

      Supports multiple content types like text, forms, files, as well as processing capabilities for audio and video streams.

      Open Ecosystem

      Community-driven open ecosystem supporting frameworks like LangGraph, CrewAI, ADK with tools & examples.

      • English
      • 中文

      Model Context Protocol 文档 MCP 中文站 Agent2Agent 文档 A2A 中文站 Agent Network Protocol 文档 ChatGPT 中文
      Powered by Agent2Agent.Info
      © 2025 Agent2Agent.Info.[email protected]