mcp-proxy-for-aws

What it is and what it does

MCP Proxy for AWS solves the authentication gap between standard MCP clients and AWS-hosted MCP servers. The Model Context Protocol (MCP) specification supports OAuth, but AWS MCP servers use AWS IAM authentication (SigV4). This package bridges that gap by automatically signing requests with your local AWS credentials, eliminating the need to build custom SigV4 logic.

You can use it two ways: as a command-line proxy that runs between your MCP client (Claude Desktop, Kiro CLI) and an AWS MCP server, or as a Python library to integrate AWS IAM-secured MCP servers directly into AI agent applications built with LangChain, LlamaIndex, or Strands Agents. It depends on boto3 and botocore for AWS credential handling and fastmcp for MCP protocol support.

Use it for:

• Connect Claude Desktop or Kiro CLI to MCP servers on AWS without manually handling SigV4 signing.
• Build Python-based AI agents using LangChain or LlamaIndex that need to call MCP servers secured with AWS IAM.
• Run as a Docker container to provide a centralized MCP proxy bridge for multiple clients accessing AWS-hosted MCP services.
• Integrate Strands Agents or other frameworks with AWS MCP servers while delegating credential management to local AWS profiles.
• Switch between multiple AWS profiles at runtime to access different MCP servers or environments without restarting.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Bridges MCP clients to AWS-hosted MCP servers using AWS IAM authentication (SigV4), working as either a lightweight proxy or a Python library for AI agent frameworks.

Yes. The package solves a real problem—AWS IAM authentication for MCP—with low install friction, active maintenance, permissive licensing, and no known vulnerabilities. It is well-positioned for teams building AI agents or using MCP clients that need to reach AWS-hosted MCP servers. Install if you need SigV4-authenticated MCP access; skip if you only use OAuth-based MCP servers or do not use MCP.

Install

mcp-proxy-for-aws on PyPI

pip install mcp-proxy-for-aws

uv add mcp-proxy-for-aws

poetry add mcp-proxy-for-aws

Installing mcp-proxy-for-aws

Before you install

Low install friction with a pure Python wheel. Active maintenance—last commit 2026-08-12, release 25 days old. Supports Python 3.10 through 3.14. Runtime depends on boto3, botocore, and fastmcp.

License in practice

Apache-2.0 permissive license allows commercial use, modification, and distribution with minimal restrictions; suitable for most production and proprietary projects.

Quickstart

# Install
pip install mcp-proxy-for-aws

# Run as proxy (requires AWS credentials configured)
uvx mcp-proxy-for-aws@1.6.4 https://your-service.us-east-1.amazonaws.com/mcp

# Or use as library
from mcp_proxy_for_aws import MCPClient
client = MCPClient(endpoint='https://your-service.us-east-1.amazonaws.com/mcp')

Requires Python 3.10+, AWS credentials configured (via AWS CLI, environment variables, or IAM roles), and an MCP server endpoint URL.

Verify before relying

• Whether the library API (MCPClient or equivalent) is documented or stable for programmatic use beyond the proxy CLI.
• Performance characteristics and latency overhead of SigV4 signing for high-frequency MCP calls.
• Compatibility scope with specific LangChain, LlamaIndex, and Strands Agents versions.

Package facts

Evidence: mcp_proxy_for_aws-1.6.4-py3-none-any.whl

Tags

More Internet packages

Further reading