@frontpoint/aws-infra-mcp
v0.1.0
Published
MCP server for generating AWS CDK infrastructure code with resource analysis
Maintainers
Readme
AWS Infra MCP
MCP (Model Context Protocol) server for generating AWS CDK TypeScript infrastructure code with resource analysis.
Features
- AWS Resource Analysis: Scan existing AWS resources (EC2, RDS, Lambda, S3, DynamoDB, ECS, EKS, ECR, VPC, API Gateway)
- CDK Code Generation: Generate type-safe AWS CDK TypeScript projects
- Full Service Coverage: Support for compute, database, storage, and network resources
- Interactive Configuration: Add resources incrementally with MCP tools
- Cost Estimation: Preview estimated monthly costs (coming soon)
- Security Validation: IAM and security group validation (coming soon)
Installation
npm install -g @frontpoint/aws-infra-mcpUsage
As MCP Server
Add to your Claude Code MCP configuration:
{
"mcpServers": {
"aws-infra": {
"command": "aws-infra-mcp",
"args": ["/path/to/your/project"]
}
}
}CLI
# Start MCP server in current directory
aws-infra-mcp
# Start with specific project path
aws-infra-mcp /path/to/projectAvailable Tools
| Tool | Description |
|------|-------------|
| aws-init | Initialize project with AWS credentials verification |
| aws-analyze | Scan existing AWS resources in your account |
| aws-status | Show current configuration status |
| aws-add-compute | Add EC2, Lambda, ECS, EKS resources |
| aws-add-database | Add RDS, Aurora, DynamoDB resources |
| aws-add-storage | Add S3, ECR resources |
| aws-add-network | Add VPC, API Gateway resources |
| aws-generate-all | Generate complete CDK project |
Supported AWS Services
Compute
- EC2 Instances & Auto Scaling Groups
- Lambda Functions with event sources
- ECS Clusters, Services, Task Definitions
- EKS Clusters with Node Groups
Database
- RDS (MySQL, PostgreSQL, MariaDB)
- Aurora (MySQL, PostgreSQL, Serverless v2)
- DynamoDB Tables with GSI/LSI
Storage
- S3 Buckets with lifecycle policies
- ECR Repositories
Network
- VPC with public/private subnets
- NAT Gateways
- API Gateway (REST, HTTP, WebSocket)
Example Workflow
# 1. Initialize project
aws-init projectName: my-app region: ap-northeast-2
# 2. Analyze existing resources (optional)
aws-analyze
# 3. Add infrastructure resources
aws-add-network type: vpc name: main-vpc cidrBlock: 10.0.0.0/16
aws-add-database type: dynamodb name: users-table partitionKey: userId:S
aws-add-storage type: s3 name: my-app-assets versioning: true
aws-add-compute type: lambda name: api-handler runtime: nodejs20.x
# 4. Generate CDK project
aws-generate-all
# 5. Deploy
cd cdk-infra && npm install && npx cdk deploy --allGenerated CDK Structure
cdk-infra/
├── bin/
│ └── app.ts # CDK app entry point
├── lib/
│ └── stacks/
│ ├── network-stack.ts
│ ├── compute-stack.ts
│ ├── database-stack.ts
│ └── storage-stack.ts
├── cdk.json
├── package.json
└── tsconfig.jsonRequirements
- Node.js >= 18.0.0
- AWS CLI configured with credentials
- AWS CDK CLI (
npm install -g aws-cdk)
Configuration
Configuration is stored in .aws-infra-mcp/ directory:
config.json- Infrastructure configurationstate.json- Generation state and historyanalysis-cache.json- Cached AWS resource analysis
License
MIT
Author
frontpoint
