Google A2A Agent Example MCP Server

@system32miro

代码仓库

基本信息

分类

AI 与智能体

传输方式

stdio

作者

system32miro

关于 Google A2A Agent Example MCP Server

Example implementation of the Google A2A protocol with a Flask server and Python client. Includes web search integration via MCP

#google-a2a#a2a-communication#flask-server

Google A2A Agent Example MCP Server 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "google-a2a-agent": {
      "command": "python",
      "args": [
        "-m",
        "venv",
        "venv"
      ]
    }
  }
}

Google A2A Agent Example MCP Server 工具

工具列表暂未提供

当前收录暂时没有可展示的工具列表，不代表服务没有工具。请查阅项目文档了解支持的能力。

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Google A2A Agent Example?

Google A2A Agent Example implements a simple Agent-to-Agent (A2A) communication system based on Google's proposed A2A protocol. It consists of a Flask-based server that acts as an AI agent and a Python client that demonstrates agent discovery and task exchange.

How to use Google A2A Agent Example?

Clone the repository, create a virtual environment, add your OpenAI and Brave API keys to a .env file, and install dependencies. Run python server.py in one terminal and python client.py in a separate terminal to send a predefined question and receive the agent's response.

Key features of Google A2A Agent Example

Implements core Google A2A protocol endpoints (/.well-known/agent.json, /tasks/send)
Uses OpenAI's GPT-4o-mini for AI query processing
Integrates Brave Search via MCP (Model Context Protocol) for web searches
Provides a client script demonstrating agent discovery and task submission
Manages external tool interactions through the pydantic-ai library

Use cases of Google A2A Agent Example

Prototyping agent-to-agent communication with Google's A2A standard
Building an AI agent that can perform web searches via Brave Search
Demonstrating agent discovery and task exchange for developer education

常见问题

What is the Google A2A protocol?

A2A is Google's proposed standard for enabling AI agents to communicate with each other through a standardized HTTP API, using an Agent Card for discovery and a Task API for exchanging messages.

What API keys are required?

You need an OpenAI API key (for GPT-4o-mini) and a Brave API key (for Brave Search), stored in a `.env` file as `OPENAI_API_KEY` and `BRAVE_API_KEY`.

How does the architecture work?

The client discovers the agent by fetching its Agent Card, sends a task to `/tasks/send`, the server processes the query using OpenAI and optionally calls Brave Search via MCP, and returns the final response to the client.

Does this implementation cover the full A2A protocol?

No, this example only implements the essential `/tasks/send` endpoint and Agent Card discovery, not the full A2A protocol specification.

Where does data processing occur?

All AI processing and tool integration (OpenAI API calls and Brave Search requests) happen server-side. The client only sends the initial task and receives the final response.

AI 与智能体分类下的更多 MCP 服务器

strac-mcp-dlp

strac-io

Detect and redact PII, PHI, PCI and secrets — SSNs, credit cards, passports, API keys and cloud credentials — in text and in local files, using the Strac DLP API. Five tools: redact_text, detect_sensitive_data, detect_fi

CrawlForge MCP Server

mysleekdesigns

28-tool MCP server for web scraping, crawling, deep research & autonomous extraction — clean Markdown & structured JSON for Claude, Cursor & any MCP client. 1,000 free one-time credits, local-Ollama LLM support.

Reach MCP

Kanbox

Operate a real LinkedIn account from your AI agent. 52 tools for the inbox, invitations, network, profiles, Sales Navigator search in plain language, posts and engagement; six playbooks served as MCP prompts that preview

OpenZiti MCP Gateway

openziti

NetFoundry MCP Gateway enables secure, isolated access to Model Context Protocol (MCP) tools across distributed systems without exposing public endpoints. Zero trust gateway for MCP servers. Aggregate, filter, and secure

AskOne: Live Q&A and Polls

phuryn

Run free live audience Q&A and polls from your agent: start an AskOne room, add and launch a poll, approve or hide questions and mark them answered, close the room, then read the questions as a FAQ draft and get the resu