README

AWS MCP

A Model Context Protocol (MCP) server that enables AI assistants like Claude to interact with your AWS environment. This allows for natural language querying and management of your AWS resources during conversations. Think of better Amazon Q alternative.

Features

🔍 Query and modify AWS resources using natural language
☁️ Support for multiple AWS profiles and SSO authentication
🌐 Multi-region support
🔐 Secure credential handling (no credentials are exposed to external services, your local credentials are used)
🏃‍♂️ Local execution with your AWS credentials

Prerequisites

Node.js
Claude Desktop
AWS credentials configured locally (~/.aws/ directory)

Installation

Clone the repository:

git clone https://github.com/RafalWilinski/aws-mcp
cd aws-mcp

Install dependencies:

pnpm install
# or
npm install

Usage

Open Claude desktop app and go to Settings -> Developer -> Edit Config

Add the following entry to your claude_desktop_config.json:

{
  "mcpServers": {
    "aws": {
      "command": "npm", // OR pnpm
      "args": [
        "--silent",
        "--prefix",
        "/Users/<YOUR USERNAME>/aws-mcp",
        "start"
      ]
    }
  }
}

Important: Replace /Users/<YOUR USERNAME>/aws-mcp with the actual path to your project directory.

Restart Claude desktop app. You should see this:

Start by selecting an AWS profile or jump to action by asking:
- "List available AWS profiles"
- "List all EC2 instances in my account"
- "Show me S3 buckets with their sizes"
- "What Lambda functions are deployed in us-east-1?"
- "List all ECS clusters and their services"

Using with `nvm`

Build from source first and add following config:

{
  "mcpServers": {
    "aws": {
      "command": "/Users/<USERNAME>/.nvm/versions/node/v20.10.0/bin/node",
      "args": [
        "<WORKSPACE_PATH>/aws-mcp/node_modules/tsx/dist/cli.mjs",
        "<WORKSPACE_PATH>/aws-mcp/index.ts",
        "--prefix",
        "<WORKSPACE_PATH>/aws-mcp",
        "start"
      ]
    }
  }
}

Troubleshooting

To see logs:

tail -n 50 -f ~/Library/Logs/Claude/mcp-server-aws.log
# or
tail -n 50 -f ~/Library/Logs/Claude/mcp.log

Features in Development

MFA support
Cache SSO credentials to prevent from refreshing them too eagerly

aws-mcp FAQ

How does AWS MCP handle AWS credentials securely?

AWS MCP uses your local AWS credentials stored in the ~/.aws/ directory and does not expose them to any external services, ensuring secure credential handling.

Can AWS MCP support multiple AWS profiles and SSO authentication?

Yes, AWS MCP supports multiple AWS profiles and Single Sign-On (SSO) authentication to manage different AWS accounts and access methods.

Is AWS MCP limited to a single AWS region?

No, AWS MCP supports multi-region AWS operations, allowing you to query and manage resources across different AWS regions.

What are the prerequisites for running AWS MCP?

You need Node.js installed, Claude Desktop for AI interaction, and AWS credentials configured locally in the ~/.aws/ directory.

Does AWS MCP run locally or in the cloud?

AWS MCP runs locally on your machine, using your local AWS credentials for secure and private execution.

How does AWS MCP integrate with AI assistants like Claude?

AWS MCP acts as a server that exposes your AWS environment to AI assistants like Claude via the Model Context Protocol, enabling natural language queries and commands.

Can AWS MCP modify AWS resources or only query them?

AWS MCP supports both querying and modifying AWS resources through natural language commands.

Is AWS MCP compatible with other AI models besides Claude?

While designed primarily for Claude, AWS MCP follows the Model Context Protocol, making it potentially compatible with other MCP-supporting AI models like OpenAI's GPT-4 and Anthropic's Claude.