AWS Interaction Protocol Server

This implements a Model Context Protocol (MCP) endpoint designed to facilitate sophisticated interaction with your Amazon Web Services (AWS) deployment using large language models like Claude. It acts as an intelligent intermediary for executing read and write operations against your cloud assets using natural language inputs, serving as a modern replacement for tools like Amazon Q.

Core Capabilities

• Execute commands for querying and modifying AWS resources via plain English.
• Compatibility across several distinct AWS credential profiles.
• Support for operations spanning numerous geographic regions.
• Execution runs locally, leveraging your existing AWS authentication secrets.

Operational Flow

Prerequisites

• A functional installation of Python.
• The Claude Desktop application must be accessible.
• Valid AWS authentication configuration must be present in the default local location (~/.aws/).

Setup Procedure

1. Retrieve the source code repository:

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

1. Install the necessary Python server components:

bash pip install -e .

Command Line Utility Usage

The MCP server also provides direct command-line access:

bash

Display available operational instructions

python -m aws_mcp --help

Enumerate configured AWS user profiles

python -m aws_mcp list-credentials

Designate a specific AWS profile for subsequent operations

python -m aws_mcp select-profile myprofile

Execute a directive against AWS

python -m aws_mcp run-aws-code "retrieve the status of all running EC2 instances"

Request neatly structured output formatting

python -m aws_mcp --format pretty list-credentials

Note: Adjust the 'python' invocation if your system's default executable path differs.

Integration with Claude Desktop

1. In the Claude desktop client, navigate to Settings -> Developer -> Configuration Editing.

2. Inject the following configuration block into your claude_desktop_config.json file:

{ "mcpServers": { "aws-mcp": { "command": "/path/to/your/python", "args": ["/path/to/your/aws-mcp/run_aws_mcp.py"] } } }

Crucial: Substitute /path/to/your/python with the output of running which python in your shell. Crucial: Replace /path/to/your/aws-mcp/run_aws_mcp.py with the absolute file system path to run_aws_mcp.py.

1. Restart the Claude application to apply changes.

2. Begin interaction by setting context or directly issuing commands, such as:

   • "What authentication profiles are available?"
   • "Detail the EC2 instances operating within my account."
   • "Report on S3 bucket volumes."
   • "Show me Lambda functions deployed in the us-east-1 region."
   • "List all ECS deployment clusters and their associated services."

Diagnostics

To monitor real-time activity logs:

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

Alternatively, check the general MCP log:

tail -n 50 -f ~/Library/Logs/Claude/mcp.log

Future Enhancements

• [ ] Incorporate support for Multi-Factor Authentication (MFA).
• [ ] Implement credential caching for Security Token Service (STS) to minimize frequent re-authentication.

Support and Community

For inquiries, feature suggestions, or contribution discussions regarding the AWS MCP utility, please utilize these channels: - Raise a new issue on the GitHub repository. - Send correspondence to: arun.sanna@outlook.com

Consult CONTRIBUTING.md for guidance on how to participate in project development.

AWS Operation Backend

AWS MCP facilitates dynamic execution by interfacing directly with the boto3 SDK. Its capabilities include:

1. Interpreting instructions provided in unstructured text or code-like formats to ascertain the required AWS service and specific API call.
2. Automatically invoking the necessary boto3 routine.
3. Presenting results in an easily digestible tabular format whenever feasible.
4. Ensuring robust error handling during operation execution.

Recognized Services

The internal parser is equipped to identify and interact with the comprehensive suite of services exposed by boto3, including but not limited to:

• Simple Storage Service (S3)
• Elastic Compute Cloud (EC2)
• AWS Lambda
• Identity and Access Management (IAM)
• DynamoDB
• Relational Database Service (RDS)
• CloudFormation
• CloudWatch
• Simple Notification Service (SNS)
• Simple Queue Service (SQS)
• Key Management Service (KMS)
• Secrets Manager
• Systems Manager (SSM)
• Route 53
• CloudFront
• Elastic Beanstalk
• API Gateway
• Elastic Container Service (ECS)
• Security Token Service (STS)
• Amazon Cognito
• CodeBuild
• CodeCommit
• CodePipeline
• AWS Organizations
• GuardDuty
• ...and numerous others!