AWS MCP Servers for Claude Code: Complete Setup Guide

Last Updated: July 11, 2025 | Version 2.0

This guide provides detailed setup instructions for all 45+ AWS MCP servers specifically for Claude Code. Each server includes claude mcp add commands, configuration options, and practical examples.

📋 Prerequisites:

Claude Code installed (claude.ai/code)

AWS CLI configured with valid credentials

Basic understanding of AWS services

💡 Pro Tip: All commands use the -s project flag to create a .mcp.json file in your project root for version control.

ℹ️ Note: This guide is specific to Claude Code's CLI. Using a different MCP client? We're planning a general MCP configuration guide with raw JSON configs - stay tuned!

Essential Setup

Setting Environment Variables

Before installing any MCP servers, set up your environment:

# Core AWS settings
export AWS_PROFILE="your-aws-profile"
export AWS_REGION="us-east-1"
export MCP_LOG_LEVEL="ERROR"
 
# Optional: Add to your shell profile for persistence
echo 'export AWS_PROFILE="your-aws-profile"' >> ~/.bashrc
echo 'export AWS_REGION="us-east-1"' >> ~/.bashrc
echo 'export MCP_LOG_LEVEL="ERROR"' >> ~/.bashrc

Installing the Core Server (Required First!)

⚠️ CRITICAL: Always install the Core MCP Server before any other AWS MCP servers.

claude mcp add awslabs.core-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.core-mcp-server@latest

Core Infrastructure

1. AWS Core MCP Server ⭐ REQUIRED

Purpose: Orchestrates and manages all other AWS MCP servers.

claude mcp add awslabs.core-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.core-mcp-server@latest

Features:

Automatic MCP server management
Planning and guidance for AWS operations
Centralized configuration
Server health monitoring

2. AWS Documentation MCP Server

Purpose: Access AWS documentation and best practices directly in Claude.

claude mcp add awslabs.aws-documentation-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_DOCUMENTATION_PARTITION=aws \
  -- uvx awslabs.aws-documentation-mcp-server@latest

Configuration Options:

AWS_DOCUMENTATION_PARTITION: Set to aws-cn for China region docs

Example Usage:

"Show me the documentation for DynamoDB TTL"
"What are the best practices for S3 bucket policies?"

3. AWS CDK MCP Server

Purpose: Infrastructure as Code with AWS CDK, including security scanning.

claude mcp add awslabs.cdk-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.cdk-mcp-server@latest

Prerequisites:

npm install -g aws-cdk

Features:

CDK best practices enforcement
CDK Nag security scanning
AWS Solutions Constructs patterns
GenAI CDK constructs

4. AWS Terraform MCP Server

Purpose: Terraform infrastructure management with security compliance.

claude mcp add awslabs.terraform-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.terraform-mcp-server@latest

Features:

Checkov security scanning
AWS provider documentation
Terraform workflow automation
Module recommendations

5. CloudFormation MCP Server

Purpose: Direct AWS resource management via CloudFormation.

claude mcp add awslabs.cfn-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.cfn-mcp-server@latest

Features:

Stack creation/update/deletion
Resource drift detection
Change set preview
Stack policy management

AI & Machine Learning

6. Amazon Nova Canvas MCP Server

Purpose: AI-powered image generation using Amazon Nova.

claude mcp add awslabs.nova-canvas-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.nova-canvas-mcp-server@latest

Features:

Text-to-image generation
Color-guided creation
Workspace integration
Multiple output formats

7. Amazon Bedrock Knowledge Base Retrieval

Purpose: Query Bedrock knowledge bases with natural language.

claude mcp add awslabs.bedrock-kb-retrieval-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.bedrock-kb-retrieval-mcp-server@latest

Features:

Knowledge base discovery
Natural language queries
Source filtering
Result reranking (us-east-1 only)

8. Amazon Q MCP Server

Purpose: Enterprise search across Q Business applications.

claude mcp add awslabs.q-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.q-mcp-server@latest

Features:

OIDC authentication
Cross-data source search
IAM role assumption
Relevance ranking

9. Amazon Kendra Index MCP Server

Purpose: Intelligent search across structured and unstructured data.

claude mcp add awslabs.kendra-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -e KENDRA_INDEX_ID="your-kendra-index-id" \
  -- uvx awslabs.kendra-mcp-server@latest

Required Configuration:

KENDRA_INDEX_ID: Your Kendra index identifier

10. Bedrock Data Automation MCP Server

Purpose: Automate data processing workflows with AI.

claude mcp add awslabs.bedrock-data-automation-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.bedrock-data-automation-mcp-server@latest

Development Tools

11. Cost Analysis MCP Server

Purpose: Analyze and optimize AWS costs before deployment.

claude mcp add awslabs.cost-analysis-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -- uvx awslabs.cost-analysis-mcp-server@latest

Features:

Real-time cost analysis
Predictive cost modeling
Resource optimization suggestions
Budget tracking

12. AWS Diagram MCP Server

Purpose: Generate architecture diagrams programmatically.

claude mcp add awslabs.aws-diagram-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.aws-diagram-mcp-server@latest

Features:

AWS architecture diagrams
Sequence diagrams
Flow charts
Custom styling

13. Frontend MCP Server

Purpose: Build React applications with AWS Amplify integration.

claude mcp add awslabs.frontend-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.frontend-mcp-server@latest

Features:

React component generation
Amplify auth integration
Theme customization
Nova Canvas integration for assets

14. Code Documentation Generation MCP Server

Purpose: Automatically generate comprehensive documentation.

claude mcp add awslabs.code-doc-gen-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.code-doc-gen-mcp-server@latest

Features:

README generation
API documentation
Architecture diagrams
Interactive workflows

15. Git Repo Research MCP Server

Purpose: Semantic search across codebases.

claude mcp add awslabs.git-repo-research-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -e GITHUB_TOKEN="your-github-token" \
  -- uvx awslabs.git-repo-research-mcp-server@latest

Required Configuration:

GITHUB_TOKEN: For accessing private repositories

Compute & Serverless

16. AWS Lambda Tool MCP Server

Purpose: Execute Lambda functions as MCP tools.

claude mcp add awslabs.lambda-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.lambda-mcp-server@latest

Optional Filters:

# Filter by prefix
-e FUNCTION_PREFIX="prod-"
 
# Filter by specific functions
-e FUNCTION_LIST="func1,func2,func3"
 
# Filter by tags
-e FUNCTION_TAG_KEY="Environment"
-e FUNCTION_TAG_VALUE="Production"

17. AWS Serverless MCP Server

Purpose: Complete serverless application development.

# Read-only mode (default)
claude mcp add awslabs.aws-serverless-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.aws-serverless-mcp-server@latest
 
# With write permissions
claude mcp add awslabs.aws-serverless-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.aws-serverless-mcp-server@latest \
  --allow-write --allow-sensitive-data-access

Features:

SAM template generation
API Gateway integration
Event source mapping
Deployment automation

Containers & Orchestration

18. Amazon EKS MCP Server

Purpose: Kubernetes cluster management on AWS.

# Read-only mode (recommended for production)
claude mcp add awslabs.eks-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.eks-mcp-server@latest
 
# With write permissions
claude mcp add awslabs.eks-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.eks-mcp-server@latest \
  --allow-write --allow-sensitive-data-access

Important: Requires EKS Access Entry or cluster creator permissions.

19. Amazon ECS MCP Server

Purpose: Container orchestration with ECS.

claude mcp add awslabs.ecs-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.ecs-mcp-server@latest

20. AWS Batch MCP Server

Purpose: Batch computing job management.

claude mcp add awslabs.batch-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.batch-mcp-server@latest

Database Services

21. Amazon DynamoDB MCP Server

Purpose: NoSQL database operations with natural language.

# Read-only mode (recommended)
claude mcp add awslabs.dynamodb-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.dynamodb-mcp-server@latest
 
# With write permissions (use cautiously)
claude mcp add awslabs.dynamodb-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -e DDB-MCP-READONLY=false \
  -- uvx awslabs.dynamodb-mcp-server@latest

22. Amazon Aurora Postgres MCP Server

Purpose: Natural language SQL for Aurora PostgreSQL.

claude mcp add awslabs.postgres-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.postgres-mcp-server@latest \
  --resource_arn "your-cluster-arn" \
  --secret_arn "your-secret-arn" \
  --database "your-database" \
  --region $AWS_REGION \
  --readonly True

23. Amazon DocumentDB MCP Server

Purpose: MongoDB-compatible document database operations.

claude mcp add awslabs.documentdb-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -e DOCDB_CLUSTER_ENDPOINT="your-cluster-endpoint" \
  -e DOCDB_USERNAME="your-username" \
  -e DOCDB_PASSWORD_SECRET_ID="your-secret-id" \
  -- uvx awslabs.documentdb-mcp-server@latest

24. Amazon ElastiCache Valkey MCP Server

Purpose: Redis-compatible in-memory data store.

claude mcp add awslabs.valkey-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e VALKEY_HOST="your-endpoint" \
  -e VALKEY_PORT="6379" \
  -e VALKEY_PASSWORD="your-password" \
  -e VALKEY_SSL=true \
  -- uvx awslabs.valkey-mcp-server@latest

25. Amazon ElastiCache Memcached MCP Server

Purpose: High-performance caching with Memcached.

claude mcp add awslabs.memcached-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e MEMCACHED_ENDPOINT="your-endpoint" \
  -e MEMCACHED_PORT="11211" \
  -e MEMCACHED_USE_TLS=true \
  -- uvx awslabs.memcached-mcp-server@latest

26. Amazon MemoryDB MCP Server

Purpose: Redis-compatible durable in-memory database.

claude mcp add awslabs.memorydb-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.memorydb-mcp-server@latest

27. Amazon ElastiCache MCP Server

Purpose: Manage ElastiCache clusters and configurations.

claude mcp add awslabs.elasticache-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.elasticache-mcp-server@latest

28. Amazon RDS MCP Server

Purpose: Relational database instance management.

claude mcp add awslabs.rds-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.rds-mcp-server@latest

29. Amazon Timestream for InfluxDB MCP Server

Purpose: Time-series database operations.

claude mcp add awslabs.timestream-influxdb-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.timestream-influxdb-mcp-server@latest

Security & Identity

30. AWS IAM MCP Server

Purpose: Identity and access management with security best practices.

# Read-only mode (strongly recommended)
claude mcp add awslabs.iam-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.iam-mcp-server@latest --readonly
 
# Full access (use with extreme caution)
claude mcp add awslabs.iam-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.iam-mcp-server@latest

31. AWS Secrets Manager MCP Server

Purpose: Secure credential and secret management.

claude mcp add awslabs.secrets-manager-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.secrets-manager-mcp-server@latest

32. AWS Systems Manager MCP Server

Purpose: Parameter Store and configuration management.

claude mcp add awslabs.ssm-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.ssm-mcp-server@latest

Monitoring & Operations

33. Amazon CloudWatch Logs MCP Server

Purpose: Log analysis and insights with natural language.

claude mcp add awslabs.cloudwatch-logs-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.cloudwatch-logs-mcp-server@latest

34. Amazon CloudWatch Metrics MCP Server

Purpose: Metric retrieval and custom metric publishing.

claude mcp add awslabs.cloudwatch-metrics-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.cloudwatch-metrics-mcp-server@latest

35. Amazon Prometheus MCP Server

Purpose: Managed Prometheus for metrics and monitoring.

claude mcp add awslabs.prometheus-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -e WORKSPACE_ID="your-workspace-id" \
  -- uvx awslabs.prometheus-mcp-server@latest

36. AWS Health MCP Server

Purpose: Monitor service health and events.

claude mcp add awslabs.health-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.health-mcp-server@latest

37. OpenAPI MCP Server

Purpose: Dynamic tool generation from OpenAPI specs.

claude mcp add awslabs.openapi-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.openapi-mcp-server@latest

Integration & Messaging

38. AWS Location Service MCP Server

Purpose: Geospatial operations and mapping.

claude mcp add awslabs.aws-location-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -e PLACE_INDEX_NAME="your-place-index" \
  -e ROUTE_CALCULATOR_NAME="your-route-calculator" \
  -- uvx awslabs.aws-location-mcp-server@latest

39. Amazon SNS/SQS MCP Server

Purpose: Message queuing and pub/sub operations.

claude mcp add awslabs.sns-sqs-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.sns-sqs-mcp-server@latest

40. Amazon EventBridge MCP Server

Purpose: Event-driven architecture management.

claude mcp add awslabs.eventbridge-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.eventbridge-mcp-server@latest

41. AWS Step Functions MCP Server

Purpose: Workflow orchestration and state machines.

claude mcp add awslabs.stepfunctions-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.stepfunctions-mcp-server@latest

42. Amazon API Gateway MCP Server

Purpose: API creation and management.

claude mcp add awslabs.apigateway-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.apigateway-mcp-server@latest

43. Amazon Rekognition MCP Server

Purpose: Computer vision and image analysis.

claude mcp add awslabs.rekognition-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.rekognition-mcp-server@latest

44. AWS HealthOmics MCP Server

Purpose: Genomics and healthcare data management.

claude mcp add awslabs.healthomics-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -e AWS_PROFILE=$AWS_PROFILE \
  -e AWS_REGION=$AWS_REGION \
  -- uvx awslabs.healthomics-mcp-server@latest

45. Synthetic Data MCP Server

Purpose: Generate test data for development.

claude mcp add awslabs.synthetic-data-mcp-server -s project \
  -e FASTMCP_LOG_LEVEL=$MCP_LOG_LEVEL \
  -- uvx awslabs.synthetic-data-mcp-server@latest

Best Practices

1. Security First

Always use read-only modes in production:

# Good: IAM in read-only mode
-- uvx awslabs.iam-mcp-server@latest --readonly
 
# Good: DynamoDB read-only
-e DDB-MCP-READONLY=true
 
# Good: Serverless without write
-- uvx awslabs.aws-serverless-mcp-server@latest

2. Environment-Specific Profiles

Create separate AWS profiles for different environments:

# ~/.aws/config
[profile dev-mcp]
region = us-east-1
output = json
 
[profile staging-mcp-readonly]
region = us-east-1
output = json
 
[profile prod-mcp-readonly]
region = us-east-1
output = json

3. Project-Level Configuration

Always use -s project to create .mcp.json:

# This creates .mcp.json in your project root
claude mcp add [server] -s project ...
 
# Version control your configuration
git add .mcp.json
git commit -m "Add MCP server configuration"

4. Logging Configuration

Adjust log levels based on your needs:

# Production (minimal logs)
export MCP_LOG_LEVEL="ERROR"
 
# Development (more verbose)
export MCP_LOG_LEVEL="INFO"
 
# Debugging (full logs)
export MCP_LOG_LEVEL="DEBUG"

5. Server Combinations

Effective server combinations for common use cases:

Full-Stack Development:

Core + Frontend + Serverless + DynamoDB
awslabs.core-mcp-server
awslabs.frontend-mcp-server
awslabs.aws-serverless-mcp-server
awslabs.dynamodb-mcp-server

Infrastructure Management:

Core + CDK + Cost Analysis + IAM
awslabs.core-mcp-server
awslabs.cdk-mcp-server
awslabs.cost-analysis-mcp-server
awslabs.iam-mcp-server --readonly

Monitoring & Operations:

Core + CloudWatch + Health + Cost
awslabs.core-mcp-server
awslabs.cloudwatch-logs-mcp-server
awslabs.cloudwatch-metrics-mcp-server
awslabs.health-mcp-server
awslabs.cost-analysis-mcp-server

Troubleshooting

Common Issues

1. "Cannot find server" error

# Clear the cache and reinstall
uv cache clean awslabs.[server-name]
claude mcp remove awslabs.[server-name]
claude mcp add awslabs.[server-name] -s project ...

2. Authentication failures

# Verify AWS credentials
aws sts get-caller-identity
 
# Check profile
echo $AWS_PROFILE
 
# Test with AWS CLI
aws s3 ls

3. Permission denied errors

# Check IAM permissions for the service
aws iam get-user-policy --user-name $USER --policy-name YourPolicy
 
# Use read-only modes
-e DDB-MCP-READONLY=true
-- uvx awslabs.iam-mcp-server@latest --readonly

4. Server not responding

# Check logs
export MCP_LOG_LEVEL="DEBUG"
# Re-run your command
 
# Check .mcp.json for conflicts
cat .mcp.json

Getting Help

Documentation: https://awslabs.github.io/mcp/
GitHub Issues: https://github.com/awslabs/mcp/issues
AWS Forums: Search for "MCP servers" in AWS forums

Next Steps

Start with Core + 2-3 servers relevant to your use case
Test in development before production use
Use read-only modes until comfortable
Document your setup in your project README
Share configurations with your team via .mcp.json

Remember: MCP servers amplify your AWS knowledge—they don't replace it. Use them as productivity tools while maintaining oversight of the infrastructure they help you create.

This guide is maintained by Sati Technology, Inc. For updates and corrections, please contact us or submit a PR to our documentation repository.