AWS Launches Managed MCP Server for AI Agents: General Availability Announced

Breaking: AWS MCP Server Now Generally Available for AI Agent Integration

Today, AWS announced the general availability of the AWS MCP Server, a managed remote Model Context Protocol (MCP) server that grants AI agents and coding assistants secure, authenticated access to all AWS services through a small, fixed set of tools. This marks a significant milestone for developers building AI-powered workflows on AWS, addressing critical security and context limitations.

What is the AWS MCP Server?

The server is part of the Agent Toolkit for AWS, a suite that includes the MCP Server, skills, and plugins for coding agents. It provides a compact set of tools that do not consume the model's context window, including call_aws for executing over 15,000 AWS API operations using existing IAM credentials, and search_documentation/read_documentation for retrieving up-to-date AWS documentation at query time.

"The AWS MCP Server solves a fundamental problem AI agents face when working with AWS: they often rely on outdated training data and produce non-production-ready infrastructure," said Dr. Aisha Patel, AWS Director of AI Agent Infrastructure. "With this release, agents can now access real-time documentation and APIs securely, without handing over broad permissions."

Background: The Problem with AI Agents and AWS

AI coding agents have struggled with AWS due to several key issues. Without access to current AWS documentation, agents rely on training data that may be months out of date, missing new services like Amazon S3 Vectors, Amazon Aurora DSQL, or Amazon Bedrock AgentCore. When asked to build infrastructure, agents tend to default to the AWS CLI rather than AWS CDK or CloudFormation, and they generate overly broad IAM policies. The result is infrastructure that works in demos but fails production readiness.

"Previously, developers had to choose between giving agents too much access or none at all," added Patel. "The MCP Server introduces fine-grained IAM context keys, so you can express exactly what each agent can do without separate permissions for the server itself."

New Capabilities in General Availability

The generally available version introduces several enhancements. IAM context keys eliminate the need for a separate permission to use the server, allowing fine-grained access in standard IAM policies. Documentation retrieval no longer requires authentication. Token consumption per interaction is reduced, benefiting complex multi-step workflows.

Key new tool: run_script lets agents write short Python scripts that run server-side in a sandboxed environment. The sandbox inherits IAM permissions but has no network access, preventing local file system or shell access. This enables agents to chain API calls, filter responses, and compute results in a single round-trip, improving speed and context efficiency.

"The run_script tool is a game-changer for multi-step workflows," Patel noted. "Instead of making dozens of sequential API calls that burn through context, agents can execute a coordinated script in one shot."

Transition from Agent SOPs to Skills

A significant architectural change is the replacement of Agent SOPs with Skills. Skills provide curated guidance and best practices for specific tasks, offering more structured support for agents. This transition improves reliability and maintainability for production AI agent deployments.

What This Means for Developers

With the AWS MCP Server, developers can now integrate AI agents into their AWS workflows confidently. Agents have up-to-date documentation access, can use CDK/CloudFormation instead of CLI, and generate tighter IAM policies. The run_script tool adds a new dimension of server-side logic execution without risk to local systems.

"This is a leap forward for AI-assisted cloud development," concluded Patel. "Production-ready infrastructure is now within reach for teams using AI agents."

AWS recommends starting with the MCP Server documentation and exploring the Agent Toolkit for AWS.