Introducing the Amazon Bedrock AgentCore MCP Server: Revolutionizing AI Agent Development

Unlocking New Possibilities in AI Development

Key Features of the AgentCore MCP Server

A Layered Approach to AI Agent Development

Getting Started: Installation Guide

Harnessing the Power of the AgentCore MCP Server for Efficient Development

Best Practices: Cleaning Up After Development

Conclusion: Enhance Your Workflow with the AgentCore MCP Server

Meet the Authors: Experts Behind the Innovation

Unleashing Innovation: Introducing Amazon Bedrock AgentCore MCP Server

Today, we’re thrilled to unveil the Amazon Bedrock AgentCore Model Context Protocol (MCP) Server, a game-changing tool designed to streamline the development of AI agents. With integrated runtime support, gateway integration, identity management, and agent memory, the AgentCore MCP Server is crafted to accelerate component creation compatible with Bedrock AgentCore. Whether you’re in rapid prototyping, building production AI solutions, or scaling your enterprise’s agent infrastructure, this is the solution you’ve been waiting for.

Transforming the Development Landscape

Agentic Integrated Development Environments (IDEs)—including Kiro, Claude Code, GitHub Copilot, and Cursor—paired with sophisticated MCP servers, are redefining how developers create AI agents. Traditionally, the learning curve associated with Bedrock AgentCore services, integrating Runtime and Tools Gateway, managing security configurations, and deploying to production took considerable time and effort. Now, using conversational commands with your coding assistant, these complex tasks can be accomplished in just minutes.

Introducing the AgentCore MCP Server

In this post, we will dive deeper into the capabilities of the AgentCore MCP Server and offer you a guide on how to get started with installation.

AgentCore MCP Server Capabilities

The AgentCore MCP Server introduces a revolutionary agentic development experience on AWS. It provides specialized tools that automate the entire agent lifecycle, minimize the learning curve, and eliminate development friction that often hinders innovation. Here’s a look at what the AgentCore MCP Server can do:

1. Agent Transformation: The server helps transition agents for AgentCore Runtime integration by guiding your coding assistant on necessary functionality changes—such as library imports and updating dependencies—all while retaining your existing agent logic and Strands Agent features.

2. Automated Environment Provisioning: It streamlines the setup process through your coding assistant, automating the installation of dependencies, configuring AWS credentials, defining roles, setting up ECR repositories, and creating configuration files.

3. Seamless Tool Integration: Simplifying communication between agents and tools within a cloud environment is a grace of the AgentCore Gateway.

4. User-Friendly Testing: It enables simple agent invocation and testing through natural language commands, making it easy to verify workflows.

A Layered Approach to Development

To maximize the benefits of the AgentCore MCP Server, we suggest a layered architecture for comprehensive AI agent development support:

• Layer 1: Agentic IDE or Client: Start with a natural language interface like Kiro or Cursor for basic tasks.

• Layer 2: AWS Service Documentation: Install the AWS Documentation MCP Server for thorough AWS service documentation.

• Layer 3: Framework Documentation: Get additional insight by installing relevant MCP servers for frameworks like Strands or LangGraph.

• Layer 4: SDK Documentation: Combine information across Agent Framework SDKs and Bedrock AgentCore SDK.

• Layer 5: Steering Files: Utilize task-specific guidance to facilitate complex workflows.

This layered approach provides robust context, allowing your coding assistant to manage everything from basic AWS operations to advanced agent transformations.

Getting Started with Installation

Starting with the Amazon Bedrock AgentCore MCP server is a breeze. You can jump right in using the one-click install available on our GitHub repository. The installation process involves editing the mcp.json file to integrate with your chosen IDE.

Here’s a basic structure to include in your mcp.json:

{
  "mcpServers": {
    "awslabs.amazon-bedrock-agentcore-mcp-server": {
      "command": "uvx",
      "args": ["awslabs.amazon-bedrock-agentcore-mcp-server@latest"],
      "env": {
        "FASTMCP_LOG_LEVEL": "ERROR"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

Examples of Agent Development

Let’s explore a typical agent development lifecycle:

1. Setup: Use local tools or deploy the MCP server on AgentCore Runtime.
2. Code Preparation: Utilize frameworks like Strands or LangGraph to prepare your agent.
3. Transformation: Request your agent to transform the code for AgentCore Runtime compatibility.
4. Deployment: Deploy your agent to Runtime and test it efficiently with a single command.
5. Integration: Modify the code for utilizing the deployed Gateway MCP server, and deploy a new version with enhanced functionality.

Clean Up

If you ever need to uninstall the MCP server, consult the MCP documentation for your specific IDE to follow the correct procedures.

Conclusion

In this post, we’ve outlined how the AgentCore MCP Server is poised to accelerate your development workflows, making the creation of AI agents easier and faster than ever. We encourage you to dive into our GitHub repository and leverage the available resources to optimize your development process.

Explore the potential of the AgentCore MCP Server, and please share your feedback through our GitHub issues page to help us enhance your experience!

About the Authors

Shreyas Subramanian: A Principal Data Scientist at AWS, Shreyas specializes in Generative AI to tackle business challenges. With expertise in optimization and deep learning, he has authored bestselling books and numerous research papers.

Primo Mu: As a Software Development Engineer, Primo builds the foundational systems behind AWS products like Kiro and Q Dev CLI, focusing on scalable frameworks and robust architectures for intelligent applications.

Explore the future of AI agent development with the Amazon Bedrock AgentCore MCP Server—where innovation meets efficiency!