---
id: bedrock-agentcore
version: "1.21.0"
license: Apache-2.0
license_treatment: permissive
maintenance: active
---

# bedrock-agentcore — An SDK for using Bedrock AgentCore

License: permissive · Maintenance: active · Downloads: 8.3M/mo

## What it is and what it does

Bedrock AgentCore SDK is a Python wrapper that lets you deploy AI agents to AWS Bedrock AgentCore—a managed platform for running agents at scale without provisioning servers. It abstracts away infrastructure concerns and provides a decorator-based entrypoint pattern compatible with any agent framework (Strands, LangGraph, CrewAI, Autogen, or custom code). The SDK handles session isolation, authentication, observability via OpenTelemetry, and persistent memory across agent invocations.

The package ships with support for two agent communication protocols: AG-UI (Server-Sent Events and WebSocket) and A2A (Agent-to-Agent), both installable as optional extras. It depends on boto3 for AWS API calls, pydantic for validation, starlette and uvicorn for the local development server, and websockets for real-time communication. The project is actively maintained but still in alpha, so expect API refinements.

Use it for:

- Deploy a local Strands or LangGraph agent to Bedrock AgentCore without writing CloudFormation or CDK.
- Build a multi-turn conversational agent with persistent memory and session isolation across API calls.
- Integrate third-party APIs as agent tools via the Gateway service and MCP protocol.
- Run untrusted agent code safely in a sandboxed Code Interpreter environment.
- Monitor agent behavior and performance with built-in OpenTelemetry tracing and observability.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Deploy AI agents to AWS Bedrock AgentCore with a Python SDK that handles runtime, memory, observability, and authentication without managing infrastructure.

Yes, if you are deploying agents to AWS and want to avoid infrastructure management. The low install friction, active maintenance, permissive license, and zero vulnerabilities make it a solid choice. However, the alpha status means APIs may change—pin the version and monitor releases. Not suitable if you need a stable, production-hardened API or are deploying outside AWS.

## Install

pip install bedrock-agentcore
uv add bedrock-agentcore
poetry add bedrock-agentcore

## Installing bedrock-agentcore

Before you install: Low install friction with a pure Python wheel and standard dependencies. Actively maintained with recent releases; repo shows 750 stars and last commit on 2026-08-11. Still in alpha status (Development Status :: 3), so expect API changes.

License in practice: Apache 2.0 permissive license allows commercial use, modification, and distribution with minimal restrictions—suitable for production deployments.

Quickstart:

pip install bedrock-agentcore

from bedrock_agentcore import BedrockAgentCoreApp
app = BedrockAgentCoreApp()

@app.entrypoint
async def handler(request):
    prompt = request.get("prompt")
    agent = Agent()  # your agent framework
    async for event in agent.stream_async(prompt):
        yield event

app.run()

Requires Python 3.10 or later; AWS credentials and Bedrock AgentCore access configured via boto3.

Verify before relying:

- Whether AG-UI and A2A protocol extras are stable or subject to breaking changes in alpha releases.
- Performance characteristics and cost model for Bedrock AgentCore runtime execution.
- Compatibility guarantees with specific agent frameworks (Strands, LangGraph, CrewAI, Autogen) across SDK versions.

## Package facts

- License: Apache-2.0 (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 8.3M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

bedrock agent deployment sdk, serverless ai agent hosting, agent framework integration aws, ag-ui protocol python, a2a agent protocol support, aws-bedrock, agent-framework, serverless-deployment

[View on SkillFed](https://skillfed.io/packages/bedrock-agentcore) · [View on PyPI](https://pypi.org/project/bedrock-agentcore/)