Mcp
VerifiedOpen source MCP servers for AWS services and operations.
What is the Mcp MCP server?
This repository hosts multiple open-source MCP servers designed to give AI assistants real-time access to AWS resources. Servers support local stdio transport and cover areas such as documentation lookup, infrastructure management, databases, and cost monitoring.
While the servers remain functional and open to contributions, AWS has introduced the Agent Toolkit for AWS as the successor with improved IAM controls and observability. The original MCP servers continue to work for development and experimentation.
Install & connect
Add this to your MCP client config. Pick your client below and copy.
{
"mcpServers": {
"mcp": {
"command": "uvx",
"args": [
"awslabs.aws-api-mcp-server"
]
}
}
}Package: awslabs.aws-api-mcp-server (pypi)
Other ways to install
uvx
uvx awslabs.aws-api-mcp-serverExample prompts
Once connected, try asking your AI client:
Security & permissions
Runs locally via stdio and requires valid AWS credentials (via environment variables or shared config) with least-privilege IAM permissions for the services it accesses.
What you can do with Mcp
Infrastructure queries
Ask an AI client to inspect running resources or suggest CloudFormation changes using the deployed MCP servers.
Documentation lookup
Retrieve the latest official AWS service documentation directly inside the conversation without leaving the editor.
Cost and operations review
Have the assistant analyze billing data or CloudWatch metrics through the provided AWS MCP servers.
How to use Mcp
- 1Clone the awslabs/mcp repository from GitHub.
- 2Install dependencies for the specific server you want to run.
- 3Configure AWS credentials locally with appropriate permissions.
- 4Add the server entry to your client's mcp.json or settings file using stdio transport.
- 5Restart the AI client and verify the connection.
Mcp: pros & cons
Pros
- +Broad coverage of AWS services across many categories
- +Fully open source with Apache 2.0 license
- +Active community and maintained by AWS Labs
- +Works with popular clients like Cursor, VS Code, and Claude Code
Cons
- –No individual tool schemas documented in the main README
- –Being superseded by the newer Agent Toolkit for AWS
- –Requires manual setup of multiple servers for full coverage
Frequently asked questions
AWS now directs users to the Agent Toolkit for AWS for production workloads while noting that the MCP servers continue to function.
User reviews
Verified reviews from the community shape this listing's rating.
Loading reviews…
Similar MCP servers
Other cloud & devops connectors worth a look.
