Framework Agents Examples¶

This guide shows how to use popular AI agent frameworks with Amazon Bedrock AgentCore Runtime.

Prerequisites¶

Before starting, ensure you've completed the QuickStart guide and have: - AWS credentials configured - AgentCore CLI installed (agentcore --help works) - A project folder with virtual environment activated

LangGraph Agent¶

LangGraph enables building stateful, multi-actor applications with LLMs.

Installation¶

pip install langchain-aws langgraph

Create the Agent¶

Create langgraph_agent.py:

from bedrock_agentcore import BedrockAgentCoreApp
from langchain_aws import ChatBedrock
from langgraph.graph import StateGraph, START, END
from langgraph.graph.message import add_messages
from typing import Annotated, TypedDict

app = BedrockAgentCoreApp()

# Define state for conversation memory
class State(TypedDict):
    messages: Annotated[list, add_messages]

# Initialize Bedrock LLM
llm = ChatBedrock(
    model_id="us.anthropic.claude-3-7-sonnet-20250219-v1:0",
    model_kwargs={"temperature": 0.7}
)

# Define the chat node that processes messages
def chat_node(state: State):
    response = llm.invoke(state["messages"])
    return {"messages": [response]}

# Build the graph
workflow = StateGraph(State)
workflow.add_node("chat", chat_node)
workflow.add_edge(START, "chat")
workflow.add_edge("chat", END)
graph = workflow.compile()

@app.entrypoint
def invoke(payload):
    user_message = payload.get("prompt", "Hello!")
    result = graph.invoke({
        "messages": [{"role": "user", "content": user_message}]
    })
    # Extract the assistant's response
    last_message = result["messages"][-1]
    return {"result": last_message.content}

if __name__ == "__main__":
    app.run()

Deploy¶

# Create requirements.txt for container
echo "langchain-aws
langgraph" > requirements.txt

# Configure and deploy
agentcore configure --entrypoint langgraph_agent.py
agentcore launch

# Test
agentcore invoke '{"prompt": "Explain LangGraph in one sentence"}'

CrewAI Agent¶

CrewAI enables building collaborative AI agent teams.

Installation¶

pip install crewai crewai-tools

Create the Agent¶

Create crewai_agent.py:

from bedrock_agentcore import BedrockAgentCoreApp
from crewai import Agent, Task, Crew, Process
import os

app = BedrockAgentCoreApp()

# Set AWS region for litellm (used by CrewAI)
os.environ["AWS_DEFAULT_REGION"] = os.environ.get("AWS_REGION", "us-west-2")

# Create an agent with specific role and capabilities
researcher = Agent(
    role="Research Assistant",
    goal="Provide helpful and accurate information",
    backstory="You are a knowledgeable research assistant with expertise in many domains",
    verbose=False,
    llm="bedrock/us.anthropic.claude-3-7-sonnet-20250219-v1:0",  # litellm format required
    max_iter=2  # Limit iterations to control costs
)

@app.entrypoint
def invoke(payload):
    user_message = payload.get("prompt", "Hello!")

    # Create a task for the agent
    task = Task(
        description=user_message,
        agent=researcher,
        expected_output="A helpful and informative response"
    )

    # Create and run the crew
    crew = Crew(
        agents=[researcher],
        tasks=[task],
        process=Process.sequential,
        verbose=False
    )

    result = crew.kickoff()
    return {"result": result.raw}

if __name__ == "__main__":
    app.run()

Deploy¶

# Create requirements.txt for container
echo "crewai
crewai-tools" > requirements.txt

# Configure and deploy
agentcore configure --entrypoint crewai_agent.py
agentcore launch

# Test
agentcore invoke '{"prompt": "What are the benefits of using CrewAI?"}'

Key Differences Between Frameworks¶

Framework	Best For	Key Features
Strands	Simple agents	Minimal setup, built-in tools, great for beginners
LangGraph	Stateful workflows	Graph-based flows, state management, complex routing
CrewAI	Multi-agent teams	Role-based agents, collaborative tasks, delegation

Common Patterns¶

Adding Tools¶

All frameworks support tools. Here's an example with Strands:

from strands import Agent, tool

@tool
def get_weather(location: str) -> str:
    """Get weather for a location."""
    return f"Weather in {location}: Sunny, 72°F"

agent = Agent(tools=[get_weather])

Error Handling¶

Always include error handling in production:

@app.entrypoint
def invoke(payload):
    try:
        user_message = payload.get("prompt", "Hello!")
        # Your agent logic here
        return {"result": response}
    except Exception as e:
        app.logger.error(f"Agent error: {e}")
        return {"error": "An error occurred processing your request"}

Using Environment Variables¶

For API keys or configuration:

import os

@app.entrypoint
def invoke(payload):
    api_key = os.environ.get("MY_API_KEY")
    # Use the API key in your agent logic

Then set the environment variable during deployment:

agentcore launch --env MY_API_KEY=your-key-here

Troubleshooting¶

Model Access Issues¶

If you see "model access denied": 1. Ensure Claude models are enabled in Bedrock console 2. Check you're using the correct model ID format 3. Verify your AWS region matches where models are enabled

CrewAI Specific Issues¶

CrewAI uses litellm, which requires: - Model format: bedrock/model-id (not just model-id) - AWS_DEFAULT_REGION environment variable set