Introducing Agent Plugins for AWS

Deploying applications to AWS typically involves researching service options, estimating costs, and writing infrastructure-as-code tasks that can slow down development workflows. Agent plugins extend coding agents with specialized skills, enabling them to handle these AWS-specific tasks directly within your development environment.

Today, we’re announcing Agent Plugins for AWS (Agent Plugins), an open source repository of agent plugins that provide coding agents with the agent skills to architect, deploy, and operate on AWS.

Today’s launch includes an initial deploy-on-aws agent plugin, which lets developers enter deploy to AWS and have their coding agent generate AWS architecture recommendations, AWS service cost estimates, and AWS infrastructure-as-code to deploy the application to AWS. We will add additional agent skills and agent plugins in the coming weeks.

Agent plugins are currently supported in Claude Code and Cursor (announced February 17). In this post, we’ll show you how to get started with Agent Plugins for AWS, explore the deploy-on-aws plugin in detail, and demonstrate how it transforms the deployment experience from hours of configuration to a simple conversation.

Why agent plugins

AI coding agents are increasingly used in software development, helping developers write, review, and deploy code more efficiently. Agent skills and the broader agent plugin packaging model are emerging as best practices for steering coding agents toward reliable outcomes without bloating model context. Instead of repeatedly pasting long AWS guidance into prompts, developers can now encode that guidance as reusable, versioned capabilities that agents invoke when relevant. This improves determinism, reduces context overhead, and makes agent behavior easier to standardize across teams. Agent plugins act as containers that package different types of expertise artifacts together. A single agent plugin can include:

• Agent skills – Structured workflows and best-practice playbooks that guide AI through complex tasks like deployment, code review, or architecture planning. Agent skills encode domain expertise as step-by-step processes.
• MCP servers – Connections to external services, data sources, and APIs. MCP servers give your assistant access to live documentation, pricing data, and other resources at runtime. Learn more about AWS MCP servers.
• Hooks – Automation and guardrails that run on developer actions. Hooks can validate changes, enforce standards, or trigger workflows automatically.
• References – Documentation, configuration defaults, and knowledge that the agent skill can consult. References make agent skills smarter without bloating the prompt.

As new types of expertise artifacts emerge in this space, they can be packaged into agent plugins, making the evolution transparent to developers.

The deploy-on-aws plugin

The initial release includes the deploy-on-aws plugin, which gives coding agents the knowledge to deploy applications to AWS with architecture recommendations, cost estimates, and infrastructure-as-code generation.

The agent plugin provides AI coding agents with a structured workflow:

1. Analyze – Scan your codebase for framework, database, and dependencies.
2. Recommend – Select optimal AWS services with concise rationale.
3. Estimate – Show projected monthly cost before committing.
4. Generate – Write CDK or CloudFormation infrastructure code.
5. Deploy – Execute your confirmation.

The initial plugin uses three MCP servers for AWS to provide comprehensive guidance:

Agent plugins for AWS in action

Here’s how Agent plugins for AWS transform the software development experience:

A full-stack developer has built an Express.js REST API on their local machine. The application connects to a PostgreSQL database and serves a React frontend. The developer is ready to deploy to AWS but isn’t sure about the best architecture.

Using Cursor or Claude Code with the deploy-on-aws plugin installed, the developer enters:

Deploy this Express app to AWS

The plugin immediately springs into action:

Step 1: Analyze – The agent scans the codebase and identifies:

• Express.js framework (Node.js 20.x)
• PostgreSQL database dependency
• Static React build in /public
• Environment variables for database connection
• Current traffic: ~1000 requests/day expected

Step 2: Recommend – Based on the analysis, the agent recommends:

• AWS App Runner for the Express.js backend (auto-scaling, managed container service)
• Amazon RDS PostgreSQL for the database (managed, automated backups)
• Amazon CloudFront + S3 for the React frontend (global CDN, cost-effective static hosting)
• AWS Secrets Manager for database credentials

Step 3: Estimate – The agent provides a cost estimate using real-time pricing data from the AWS Pricing MCP server, giving you visibility into projected monthly costs before you commit to any infrastructure.

Step 4: Generate The developer reviews the estimate and confirms. The agent generates:

• AWS CDK infrastructure code in TypeScript
• Dockerfile for the Express app
• Database migration scripts
• Environment configuration
• GitHub Actions workflow for CI/CD

Step 5: Deploy The developer reviews the generated code, makes minor adjustments to database schema, and confirms deployment. The agent:

• Provisions all AWS resources via CDK
• Builds and deploys the container to App Runner
• Creates the Amazon RDS database and runs migrations
• Uploads the React build to S3 and configures CloudFront
• Stores credentials in Secrets Manager

Within minutes, the developer’s application is live at a custom App Runner URL, with the React frontend served globally via CloudFront. The agent provides:

• Application URLs (backend and frontend)
• Database connection details
• CloudWatch dashboard links for monitoring
• Cost tracking setup

What would have taken hours of reading documentation, comparing services, and writing infrastructure code took less than 10 minutes with the deploy-on-aws plugin. Developers can now focus on building features instead of wrestling with cloud deployment complexity.

Getting started with Agent Plugins for AWS

Prerequisites

To get started, you need:

• An agent plugin compatible AI coding tool (Claude Code, Cursor, or other compatible tools)
• AWS CLI configured with appropriate credentials

Installation

Claude Code

Add the Agent Plugins for AWS marketplace to Claude Code:/plugin marketplace add awslabs/agent-plugins

Install the deploy-on-aws plugin:

/plugin install deploy-on-aws@awslabs-agent-plugins

Cursor

Cursor announced support for agent plugins on February 17. You can install the deploy-on-aws plugin directly from the Cursor Marketplace, or manually in Cursor by:

1. Open Cursor Settings
2. Navigate to Plugins, and in the search bar type aws
3. Select the plugin you want to install, and Click add to cursor, then select the scope
4. Now the plugin should appear under Plugins, installed

Learn more in the Cursor Marketplace announcement.

Skill triggers

The deploy-on-aws plugin responds to natural language requests like:

• “Deploy to AWS”
• “Host on AWS”
• “Run this on AWS”
• “AWS architecture for this app”
• “Estimate AWS cost”
• “Generate infrastructure”

Best practices for plugin-assisted development

To maximize the benefits of plugin-assisted development while maintaining security and code quality, follow these essential guidelines:

• Always review generated code before deployment (for example, against your constraints for security, cost, resilience)
• Use plugins as accelerators, not replacements for developer judgment and expertise.
• Keep plugins updated to benefit from the latest AWS best practices.
• Follow the principle of least privilege when configuring AWS credentials.
• Run security scanning tools on generated infrastructure code.

Conclusion

In this post, we showed how Agent Plugins for AWS extend coding agents with skills for deploying applications to AWS. Using the deploy-on-aws plugin, you can generate architecture recommendations, cost estimates, and infrastructure-as-code directly from your coding agent.

Beyond deployments, agent plugins can help with other AWS workflows; more agent plugins for AWS are launching soon. You can also use AWS MCP servers to give your coding agent access to specialized tools to build on AWS.

• Visit the Agent Plugins for AWS repository to install and configure your agent plugins
• Install the deploy-on-aws plugin from the Cursor Marketplace and start deploying from your editor

About the authors