A Claude Code plugin that provides an AWS Solutions Architect agent for cloud architecture guidance. While implementation-focused plugins help you write code, this one helps you make the architectural decisions that shape your AWS infrastructure: service selection, cost optimization, security review, migration strategy, and Well-Architected alignment.
| Agent | Description |
|---|---|
aws-solutions-architect |
AWS architecture advisor covering service selection, Well-Architected reviews, cost estimation, security review, migration planning, and CDK/CloudFormation guidance |
| Skill | Description |
|---|---|
/well-architected-review |
Framework review across all six Well-Architected pillars |
This plugin bundles the AWS Documentation MCP server (installed as awslabs.aws-documentation-mcp-server via uvx, registered in .mcp.json as awslabs-aws-documentation-mcp-server), giving the agent access to current AWS documentation for any service, resource type, or best practice.
No manual MCP server configuration needed.
Add the marketplace to your Claude Code project, then install the plugin:
/plugin marketplace add cpliakas/claude-code-aws-solutions-architect
/plugin install aws-solutions-architect
@agents/aws-solutions-architect Should we use DynamoDB or Aurora for our order tracking system? We have ~2,000 orders/day with complex queries for reporting.
The agent evaluates both services against your access patterns, scaling needs, cost profile, and operational maturity, then provides a recommendation with clear trade-offs.
@agents/aws-solutions-architect Can you review our architecture against the Well-Architected Framework?
The agent invokes /well-architected-review, scanning your infrastructure code and evaluating the workload across all six pillars. Produces a structured report with prioritized recommendations.
@agents/aws-solutions-architect Can you estimate the monthly cost for running 3 ECS Fargate tasks with an ALB, RDS Multi-AZ, and ElastiCache?
The agent produces a per-service cost breakdown with assumptions, hidden cost warnings, and optimization opportunities.
| Agent | Skill | |
|---|---|---|
| What | A specialist persona with domain expertise | A repeatable procedure |
| Memory | Yes, learns across sessions | No, runs the same way each time |
| Judgment | Yes, decides how to approach problems | No, follows a defined process |
The agent uses memory: project. The agent definition ships with the plugin, but each project maintains its own memory in .claude/agent-memory/. Project-specific learnings (your account structure, services, decisions) stay in project memory.
.claude/agent-memory/aws-solutions-architect/PROJECT.md # Shared project context
.claude/agent-memory/aws-solutions-architect-aws-solutions-architect/ # Agent memory
claude-code-aws-solutions-architect/
├── .claude-plugin/
│ └── marketplace.json
├── .claude/
│ └── settings.json
├── .mcp.json
├── agents/
│ └── aws-solutions-architect.md
├── benchmarks/
│ ├── run.sh
│ └── questions/
├── skills/
│ └── well-architected-review/SKILL.md
├── AGENTS.md
├── CLAUDE.md
├── README.md
└── LICENSE
MIT