Verwenden Sie ein beliebiges Agent-Framework

Sie können Open-Source-KI-Frameworks verwenden, um einen Agenten oder ein Tool zu erstellen. Dieses Thema enthält Beispiele für eine Vielzahl von Frameworks, darunter Strands Agents und Google ADK. LangGraph

Strands Agents

Das vollständige Beispiel finden Sie unterhttps://github.com/awslabs/amazon-bedrock-agentcore-samples/tree/main/03-integrations/agentic-frameworks/strands-agents.

import os
from strands import Agent
from strands_tools import file_read, file_write, editor

agent = Agent(tools=[file_read, file_write, editor])

from bedrock_agentcore.runtime import BedrockAgentCoreApp
app = BedrockAgentCoreApp()
@app.entrypoint
def agent_invocation(payload, context):
    """Handler for agent invocation"""
    user_message = payload.get("prompt", "No prompt found in input, please guide customer to create a json payload with prompt key")
    result = agent(user_message)
    print("context:\n-------\n", context)
    print("result:\n*******\n", result)
    return {"result": result.message}
app.run()

LangGraph

Das vollständige Beispiel finden Sie unterhttps://github.com/awslabs/amazon-bedrock-agentcore-samples/tree/main/03-integrations/agentic-frameworks/langgraph.

from langchain.chat_models import init_chat_model
from typing_extensions import TypedDict
from langgraph.graph import StateGraph, START
from langgraph.graph.message import add_messages
from langgraph.prebuilt import ToolNode, tools_condition
#------------------------------------------------
from bedrock_agentcore.runtime import BedrockAgentCoreApp
app = BedrockAgentCoreApp()
#------------------------------------------------

llm = init_chat_model(
    "us.anthropic.claude-3-5-haiku-20241022-v1:0",
    model_provider="bedrock_converse",
)

# Create graph
graph_builder = StateGraph(State)
...
# Add nodes and edges
...
graph = graph_builder.compile()

# Finally write your entrypoint
@app.entrypoint
def agent_invocation(payload, context):

    print("received payload")
    print(payload)

    tmp_msg = {"messages": [{"role": "user", "content": payload.get("prompt", "No prompt found in input, please guide customer as to what tools can be used")}]}
    tmp_output = graph.invoke(tmp_msg)
    print(tmp_output)

    return {"result": tmp_output['messages'][-1].content}

app.run()

Google Agent Development Kit (ADK)

Das vollständige Beispiel finden Sie unterhttps://github.com/awslabs/amazon-bedrock-agentcore-samples/tree/main/03-integrations/agentic-frameworks/adk.

from google.adk.agents import Agent
from google.adk.runners import Runner
from google.adk.sessions import InMemorySessionService
from google.adk.tools import google_search
from google.genai import types
import asyncio
import os

# adapted form https://google.github.io/adk-docs/tools/built-in-tools/#google-search

APP_NAME="google_search_agent"
USER_ID="user1234"

# Agent Definition
# Add your GEMINI_API_KEY
root_agent = Agent(
    model="gemini-2.0-flash",
    name="openai_agent",
    description="Agent to answer questions using Google Search.",
    instruction="I can answer your questions by searching the internet. Just ask me anything!",
    # google_search is a pre-built tool which allows the agent to perform Google searches.
    tools=[google_search]
)

# Session and Runner
async def setup_session_and_runner(user_id, session_id):
    session_service = InMemorySessionService()
    session = await session_service.create_session(app_name=APP_NAME, user_id=user_id, session_id=session_id)
    runner = Runner(agent=root_agent, app_name=APP_NAME, session_service=session_service)
    return session, runner

# Agent Interaction
async def call_agent_async(query, user_id, session_id):
    content = types.Content(role='user', parts=[types.Part(text=query)])
    session, runner = await setup_session_and_runner(user_id, session_id)
    events = runner.run_async(user_id=user_id, session_id=session_id, new_message=content)

    async for event in events:
        if event.is_final_response():
            final_response = event.content.parts[0].text
            print("Agent Response: ", final_response)

    return final_response

from bedrock_agentcore.runtime import BedrockAgentCoreApp
app = BedrockAgentCoreApp()

@app.entrypoint
def agent_invocation(payload, context):
    return asyncio.run(call_agent_async(payload.get("prompt", "what is Bedrock Agentcore Runtime?"), payload.get("user_id",USER_ID), context.session_id))

app.run()

OpenAI Agents SDK

Das vollständige Beispiel finden Sie unterhttps://github.com/awslabs/amazon-bedrock-agentcore-samples/tree/main/03-integrations/agentic-frameworks/openai-agents.

from agents import Agent, Runner, WebSearchTool
import logging
import asyncio
import sys

# Set up logging
logging.basicConfig(
    level=logging.DEBUG,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
    handlers=[
        logging.StreamHandler(sys.stdout)
    ]
)
logger = logging.getLogger("openai_agents")

# Configure OpenAI library logging
logging.getLogger("openai").setLevel(logging.DEBUG)

logger.debug("Initializing OpenAI agent with tools")
agent = Agent(
    name="Assistant",
    tools=[
        WebSearchTool(),
    ],
)

async def main(query=None):
    if query is None:
        query = "Which coffee shop should I go to, taking into account my preferences and the weather today in SF?"

    logger.debug(f"Running agent with query: {query}")

    try:
        logger.debug("Starting agent execution")
        result = await Runner.run(agent, query)
        logger.debug(f"Agent execution completed with result type: {type(result)}")
        return result
    except Exception as e:
        logger.error(f"Error during agent execution: {e}", exc_info=True)
        raise

# Integration with Bedrock AgentCore
from bedrock_agentcore.runtime import BedrockAgentCoreApp
app = BedrockAgentCoreApp()

@app.entrypoint
async def agent_invocation(payload, context):
    logger.debug(f"Received payload: {payload}")
    query = payload.get("prompt", "How can I help you today?")

    try:
        result = await main(query)
        logger.debug("Agent execution completed successfully")
        return {"result": result.final_output}
    except Exception as e:
        logger.error(f"Error during agent execution: {e}", exc_info=True)
        return {"result": f"Error: {str(e)}"}

# Run the app when imported
if __name__== "__main__":
    app.run()