Hello everyone!! Recently, I preparing exam as AWS Certified Solution Architect Professional. I absolutely can't lose this exam! Because if I lose this exam, I lost both certification SAP and SAA. Of cource, I'm really looking forward to going to AWS Ambassador Global Summit next week!
This post introduces my favorite recent AWS topics. Especially, Blog post titled "Unlocking the power of Model Context Protocol (MCP) on AWS" is very attention for me! Please take a look!
Collection period: 2025/06/02 Mon ~ 2025/06/08 Sun
AWS Blog
Unlocking the power of Model Context Protocol (MCP) on AWS | AWS Machine Learning Blog
AWS Introduces Model Context Protocol (MCP) for Enhanced AI Integration
AWS has published a comprehensive guide on the Model Context Protocol (MCP), an open standard that enables AI systems to communicate with external data sources and tools. Developed by Anthropic, MCP addresses the fundamental challenge of connecting language models to enterprise data beyond their training datasets.
Key Benefits:
- Solves the "M×N integration problem" by reducing custom integrations from M×N to M+N implementations
- Creates standardized interfaces between AI models and diverse data sources
- Enables secure, scalable access to AWS services like S3, DynamoDB, RDS, and CloudWatch
Architecture:
MCP uses a client-server model where MCP clients (AI applications) communicate with MCP servers (data source interfaces) through three core primitives: Tools (functions for data retrieval), Resources (contextual data), and Prompts (interaction templates).
Amazon Bedrock Integration:
The protocol integrates seamlessly with Amazon Bedrock's Converse API, enabling language models to request external information through tool calls. This creates a natural bridge between model capabilities and enterprise systems while maintaining AWS security controls.
Practical Implementation:
AWS provides a detailed example connecting MCP to Amazon Bedrock Knowledge Bases, featuring:
- Knowledge base discovery capabilities
- Natural language querying with configurable results
- Optional reranking for improved relevance
- Data source filtering for targeted searches
Enterprise Value:
For AWS customers, MCP streamlines integration between Bedrock language models and AWS services, leverages existing security mechanisms like IAM, and enables composable AI solutions following AWS architectural best practices. The implementation transforms simple data retrieval into intelligent discovery, allowing users to ask natural language questions without understanding underlying data structures. AWS plans to expand MCP capabilities with remote implementations and advanced scaling features in future releases.
Access Claude Sonnet 4 in Amazon Q Developer CLI | AWS DevOps & Developer Productivity Blog
Amazon Q Developer now supports Claude Sonnet 4 in its CLI at no additional cost, offering advanced coding capabilities with a 72.7% score on SWE-bench agentic coding. The model excels at code analysis, bug fixes, bash commands, and feature development with precise responses.
Available Models:
- Claude Sonnet 4: High-performance with balanced intelligence
- Claude Sonnet 3.7: Extended thinking capability (default)
- Claude Sonnet 3.5: High-performance intelligent model
How to Use Claude Sonnet 4:
Requires Amazon Q Developer CLI v1.11.0+. Access via:
- /model command during chat, select claude-4-sonnet
- Start new chat: q chat --model claude-4-sonnet
- Set as default: q settings chat.defaultModel claude-4-sonnet
Model Selection Priority:
- Current session selections
- User-configured preferences
- System default (Claude 3.7 Sonnet)
Key Features:
- Seamless model switching during active sessions
- Chat continuity across sessions
- Maintains previously selected models when resuming conversations
Demonstration:
The article shows Claude Sonnet 4 creating a Python to-do list app that exceeded requirements by implementing sophisticated command parsing, comprehensive error handling, object-oriented design with type hints, and automatic README documentation generation. Claude Sonnet 4 represents a significant advancement in Amazon Q Developer's capabilities, helping developers accomplish both complex and routine tasks efficiently while adapting to user preferences and maintaining workflow efficiency.
Dynamically routing requests with Amazon API Gateway routing rules | AWS Compute Blog
Amazon API Gateway now supports dynamic routing rules for custom domain names across all AWS regions. This new capability enables routing API requests based on HTTP header values, either independently or combined with URL paths, without modifying API endpoints.
Key Features:
- Dynamic Routing Rules: Route traffic based on HTTP headers, base paths, or both combinations
- Three Routing Modes: API mappings only (default), routing rules then API mappings (hybrid), and routing rules only
- Flexible Conditions: Support up to two header conditions and one base path condition per rule, with wildcard matching for headers
- Priority System: Rules evaluated by priority (1-1,000,000), with lowest numbers having highest precedence
Use Cases:
- API versioning without URL path changes
- A/B testing with specific customer segments
- Gradual rollouts and canary deployments
- Cell-based architecture routing
- Tenant-based routing
Configuration:
Routing rules have three properties: Conditions (criteria to match), Actions (what happens when matched), and Priority (evaluation order). Header matching supports exact, prefix, suffix, and contains patterns using wildcards.
Migration:
Organizations can adopt header-based routing with zero downtime using the hybrid mode, gradually transitioning from existing base path mappings to routing rules while maintaining fallback capabilities.
Observability:
Enhanced access logging provides visibility into routing decisions through context variables like $context.customDomain.routingRuleIdMatched. The feature is backward compatible, cost-effective (no additional charges), and integrates seamlessly with existing API Gateway capabilities for both public and private REST APIs.
