mcp and Lambda-MCP-Server
These are ecosystem siblings—the official AWS MCP servers provide standardized integrations with AWS services, while the Lambda-based implementation offers an alternative deployment pattern for running MCP servers in a serverless environment.
About mcp
awslabs/mcp
Official MCP Servers for AWS
Provides specialized MCP servers for AWS services including documentation access, infrastructure-as-code tools (CloudFormation, Terraform), databases, AI/ML services, and cost management—designed to integrate with AI coding assistants (Kiro, Cline, Cursor, Windsurf) and Claude Desktop via stdio transport. Enables LLM applications to access AWS context and perform authenticated API operations, with support for both local execution and remote deployment via AWS Lambda handlers.
About Lambda-MCP-Server
mikegc-aws/Lambda-MCP-Server
Creates a simple MCP tool server with "streaming" HTTP.
Implements serverless MCP tools on AWS Lambda with minimal boilerplate—decorate Python functions with `@mcp_server.tool()` and handle requests through API Gateway with Streamable HTTP transport. Features built-in DynamoDB-backed session state management for maintaining context across tool invocations, plus Bearer token authentication via API Gateway custom authorizers. Targets MCP clients supporting Streamable HTTP (like MCP Inspector and Strands Agents).
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work