Skip to content
Kitbay
MCP serverOfficial

awslabs.sagemaker-ai-mcp-server

Runs locallyApache-2.0Updated on Nov 10, 20269,904GitHub starsCan change dataNeeds an API key

An AWS Labs MCP server for Amazon SageMaker AI. At present its tools manage SageMaker HyperPod clusters orchestrated by Amazon EKS or Slurm: deploying clusters and handling nodes and their lifecycle.

What it does

An AWS Labs MCP server for Amazon SageMaker AI. At present its tools manage SageMaker HyperPod clusters orchestrated by Amazon EKS or Slurm: deploying clusters and handling nodes and their lifecycle. Unlike several other AWS Labs servers, the README documents write access and access to sensitive data as on by default in its example setup, through the --allow-write and --allow-sensitive-data-access arguments; removing them puts the server in read-only mode. It runs locally with uvx on Python 3.10 and needs the AWS CLI configured with credentials.

Capabilities

  • Deploy SageMaker HyperPod clusters
  • Manage HyperPod cluster nodes
  • Handle HyperPod cluster lifecycle operations