# AWS Advisor MCP Server

A Model Context Protocol (MCP) server that provides AWS service recommendations based on your use case descriptions.

## Features

- **suggest_aws_service**: Get AWS service recommendations by describing your use case
- **list_aws_categories**: Browse AWS services organized by category (compute, storage, database, etc.)

## Setup

1. **Install uv** (if not already installed):
```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
```

2. **Sync dependencies**:
```bash
uv sync
```

This will create a virtual environment and install all dependencies.

## Configuration for Cursor IDE

Add this server to your Cursor MCP settings:

1. Open Cursor Settings
2. Navigate to the MCP section
3. Add a new server with this configuration:

```json
{
  "mcpServers": {
    "aws-advisor": {
      "command": "uv",
      "args": [
        "--directory",
        "/Users/antonioschaffert/workspace/tony/my-dev-mcp",
        "run",
        "python",
        "src/aws_advisor_server.py"
      ]
    }
  }
}
```

Or add it to your MCP config file (usually `~/.cursor/mcp_config.json` or similar):

```json
{
  "aws-advisor": {
    "command": "uv",
    "args": [
      "--directory",
      "/Users/antonioschaffert/workspace/tony/my-dev-mcp",
      "run",
      "python",
      "src/aws_advisor_server.py"
    ]
  }
}
```

## Usage

Once configured, you can use the tools in Cursor:

### Example prompts:

- "What AWS service should I use for serverless computing?"
- "Suggest AWS services for storing images"
- "What's the best AWS service for a relational database?"
- "Show me AWS services for real-time data streaming"
- "List all AWS categories"

### Available Tools:

1. **suggest_aws_service**
   - Input: `use_case` (string) - Description of what you want to build
   - Returns: Recommended AWS services with explanations

2. **list_aws_categories**
   - No input required
   - Returns: All AWS service categories and their services

## Testing Locally

You can test the server directly:

```bash
uv run python src/aws_advisor_server.py
```

Then send MCP protocol messages via stdin (for advanced testing).

## Covered AWS Services

- **Compute**: EC2, Lambda, ECS, EKS, Fargate, Lightsail
- **Storage**: S3, EBS, EFS, Glacier, FSx
- **Database**: RDS, DynamoDB, Aurora, DocumentDB, ElastiCache, Neptune, Redshift
- **Networking**: VPC, CloudFront, Route53, API Gateway, Direct Connect, ELB
- **Analytics**: Athena, EMR, Kinesis, Glue, QuickSight
- **ML/AI**: SageMaker, Rekognition, Comprehend, Polly, Transcribe, Translate, Lex
- **Security**: IAM, Cognito, KMS, Secrets Manager, WAF, GuardDuty
- **Messaging**: SQS, SNS, EventBridge, SES
- **Monitoring**: CloudWatch, X-Ray, CloudTrail

## Requirements

- Python 3.10+
- mcp package (>=0.9.0)