The AWS MCP Servers project is an official collection of Model Context Protocol servers maintained by AWS Labs that give AI agents structured access to AWS cloud services, documentation, and infrastructure data. With over 8,500 GitHub stars, it is one of the most popular MCP server projects and represents Amazon’s official investment in the MCP ecosystem.

The collection includes multiple specialized MCP servers. The AWS Documentation MCP Server provides access to the latest AWS documentation, API references, What’s New posts, and Getting Started guides. This allows agents to look up current service documentation, resolve API questions, and stay current with AWS feature releases without relying on potentially outdated training data.

A remote managed MCP server hosted by AWS combines comprehensive AWS API support with documentation access, providing a single integration point for agents that need both operational access and reference material. This managed server handles authentication, rate limiting, and access control on the AWS side.

The servers integrate with the standard MCP client ecosystem including VS Code, Cursor, Claude Desktop, Claude Code, Codex, and other compatible tools. Configuration uses standard MCP server JSON patterns with uvx-based invocation for Python servers and npx for TypeScript servers. Environment variables control log levels, AWS partition selection, and user agent strings.

Key use cases include infrastructure planning with current AWS best practices, troubleshooting AWS service issues with up-to-date documentation context, generating Terraform or CloudFormation configurations informed by latest API schemas, cost analysis and optimization recommendations, and security posture assessment. The servers are distributed under the Apache-2.0 license and are actively maintained with frequent updates tracking AWS service changes. The project is hosted at github.com/awslabs/mcp.