The awslabs/mcp repository is AWS’s open-source suite of specialized Model Context Protocol servers, giving AI assistants structured, tool-based access to AWS services and documentation. Maintained by AWS Labs and approaching 10,000 stars on GitHub, it is the official starting point for connecting MCP clients like Claude Code, Cursor, and Amazon Q to your AWS environment.
Rather than one monolithic server, the project ships dozens of focused servers: a core server that orchestrates the others, an AWS documentation server that searches and fetches official docs, a Knowledge server for AWS guidance, and service-specific servers for areas such as Bedrock knowledge bases, Lambda, ECS, DynamoDB, S3, CloudWatch, cost analysis, infrastructure as code, and more. Each server is a small Python package installable with uvx, exposes well-scoped tools, and follows AWS security practices such as least-privilege IAM credentials supplied by your environment. A central documentation site catalogs every server with its tool list and configuration examples, and new servers are added as AWS services and MCP patterns evolve, so the suite keeps pace with how teams actually build on AWS.
Free. The MCP servers are open source under the Apache-2.0 license. You only pay for the AWS resources your assistant actually uses, under your own AWS account pricing.
Developers and platform teams working on AWS who want their AI coding assistant to read docs, inspect resources, and operate services through governed, auditable tools instead of ad-hoc shell access.