Open source MCP servers for AWS provide up-to-date documentation, guidance, and AWS-specific tooling to accelerate AI-assisted cloud development.
Configuration
View docs{
"mcpServers": {
"awslabs-mcp": {
"command": "uvx",
"args": [
"awslabs.aws-documentation-mcp-server@latest"
],
"env": {
"FASTMCP_LOG_LEVEL": "ERROR"
}
}
}
}You can leverage open source MCP servers for AWS to connect your AI tools with up-to-date AWS documentation, guidance, and cloud-native workflows. These MCP servers expose standardized capabilities that you can access from your AI assistants or IDEs to enhance cloud development, infrastructure management, and data tasks with AWS context and best practices.
You use an MCP client to connect to one or more MCP servers. Each server provides a specific capability, such as accessing official AWS documentation, guiding infrastructure-as-code decisions, or enabling AWS service interactions within your AI workflows. Start by choosing the servers that match what you build, install them as MCP tools in your development environment, and then invoke the tools from your AI assistant or IDE. When you prompt your assistant, you can reference the server by its tool name (for example, AWS Documentation MCP Server) and request the action you want, such as retrieving API references, architecture guidance, or best practices.
To install MCP servers for AWS you typically follow a straightforward flow: install the MCP client runtime (such as uv or a corresponding tool), install Python if required, and then configure MCP servers in your chosen client (Kiro, Cursor, or VS Code). Below are concrete steps you can follow to begin using MCP servers in your environment.
Prerequisites
- Ensure you have Python installed on your system
- Install the MCP client runtime (for example, uvx or uv) as directed by your chosen MCP client
- Have AWS credentials configured with the necessary permissions
Step 1: Install the MCP client runtime
- Install uv from your package manager or follow the recommended installation flow for your environment
Step 2: Install and configure MCP servers
- Choose the AWS MCP Server you want to use and add it to your MCP client configuration
- Use the provided configuration snippet or install link to enable the server
Step 3: Start using MCP servers
- Open your MCP client and verify that the AWS MCP Server appears in the Installed or Available Tools section
- Prompt your AI assistant to use the desired MCP server, for example: 'Use AWS Documentation MCP Server to fetch the latest API references'.Configuration and setup details vary by client and server. You typically provide an executable command, a list of arguments, and optional environment variables. Some servers also require a URL for remote access if they are provided as HTTP-based endpoints. When you configure a server, include any environment variables shown in the setup examples.
Security and governance are important when integrating MCP servers. Use IAM-based permissions and auditing to ensure compliant access to AWS resources. If you plan to run servers in containers or in remote environments, follow best practices for credential management and least-privilege access.
The AWS MCP Server family covers a broad range of domains, including infrastructure as code, data and analytics, AI and machine learning, developer tools, integration and messaging, cost and operations, and healthcare workloads. You can combine multiple MCP servers to compose powerful workflows, such as using the AWS Documentation MCP Server to surface up-to-date guidance while using an IaC MCP Server to guide CDK or CloudFormation tasks.
Fetches the latest AWS documentation, API references, and related materials for your AI workflows.
Guides infrastructure-as-code tasks with best practices and AWS service guidance.
Retrieves Bedrock knowledge base content with citations for use in AI responses.