agent-communication-protocol

Agent Communication Protocol (ACP)

ACP is an open protocol for agent interoperability that solves the growing challenge of connecting AI agents, applications, and humans. It enables standardized communication across different frameworks and infrastructures.

Key Features

• REST-Based Design - Straightforward HTTP endpoints (not JSON-RPC)
• Multi-Modal - Supports text, images, audio, video, and any MIME type
• Async-First - Built primarily for asynchronous communication with sync support
• Streaming - Real-time Server-Sent Events (SSE) for incremental updates
• Stateful/Stateless - Sessions for maintaining state across interactions
• Offline Discovery - Agents remain discoverable through embedded metadata
• Framework Agnostic - Works with BeeAI, LangGraph, CrewAI, LlamaIndex, and more

Installation

# Create new project with uv
uv init --python '>=3.11' my_acp_project
cd my_acp_project

# Add ACP SDK
uv add acp-sdk

# Or with pip
pip install acp-sdk

Quick Start: Creating an Agent

Basic Echo Agent

import asyncio
from collections.abc import AsyncGenerator

from acp_sdk.models import Message
from acp_sdk.server import Context, RunYield, RunYieldResume, Server

server = Server()

@server.agent()
async def echo(
    input: list[Message], context: Context
) -> AsyncGenerator[RunYield, RunYieldResume]:
    """Echoes everything"""
    for message in input:
        await asyncio.sleep(0.5)
        yield {"thought": "I should echo everything"}
        await asyncio.sleep(0.5)
        yield message

server.run()

Run with:

uv run agent.py

Server starts at http://localhost:8000.

Agent with Metadata

@server.agent(
    name="data-analyzer",
    description="Analyzes datasets and generates insights"
)
async def DataAnalyzerAgent(
    input: list[Message], context: Context
) -> AsyncGenerator[RunYield, RunYieldResume]:
    # Implementation
    yield str(response.object)

Agent Manifest