SageMaker AI MCP Server

A Model Context Protocol (MCP) server for Amazon SageMaker AI that enables AI assistants to access, work and manage SageMaker AI resources.

Features

• Managing and working with SageMaker AI endpoint resources
• Managing and working with SageMaker AI training, processing and transform jobs
• Managing and working with SageMaker AI pipelines
• CRUD operations for SageMaker AI Managed MLflow Tracking Server
• CRUD operations for SageMaker AI Domain
• Managing and working with Models and Model Cards
• CRUD operatioons for SageMaker AI Recommender Jobs
• CRUD operations for SageMaker AI Apps

Prerequisites

1. Install uv from Astral or the GitHub README
2. Install Python using uv python install 3.10
3. Set up AWS credentials with access to Amazon SageMaker AI resources.
   • You need an AWS account with Amazon SageMaker AI enabled
   • Configure AWS credentials with aws configure or environment variables
   • Ensure your IAM role/user has permissions to use Amazon SageMaker AI
4. Create a SageMaker Execution Role with the necessary permissions for SageMaker AI operations

Installation

WIP

Environment Variables

• AWS_PROFILE: AWS CLI profile to use for credentials
• AWS_REGION: AWS region to use (default: us-east-1)
• SAGEMAKER_EXECUTION_ROLE_ARN: ARN of the SageMaker execution role

AWS Authentication

The server uses the AWS profile specified in the AWS_PROFILE environment variable. If not provided, it defaults to the default credential provider chain.

"env": {
  "AWS_PROFILE": "your-aws-profile",
  "AWS_REGION": "us-east-1",
  "SAGEMAKER_EXECUTION_ROLE_ARN": "arn:aws:iam::123456789012:role/SageMakerExecutionRole"
}

Make sure the AWS profile has permissions to access Amazon SageMaker AI services. The MCP server creates a boto3 session using the specified profile to authenticate with AWS services.

Tools

List of Tools for SageMaker AI Endpoints and Endpoint Configurations

• list_endpoints_sagemaker (List all SageMaker AI Endpoints)
• list_endpoint_configs_sagemaker (List all SageMaker AI Endpoint Configurations)
• describe_endpoint_sagemaker (Describe a SageMaker AI Endpoint)
• describe_endpoint_config_sagemaker (Describe a SageMaker AI Endpoint Configuration)
• delete_endpoint_sagemaker (Delete a SageMaker AI Endpoint)
• delete_endpoint_config_sagemaker (Delete a SageMaker AI Endpoint Configuration)

List of Tools for SageMaker AI Jobs

• list_training_jobs_sagemaker (List all SageMaker AI Training Jobs)
• list_processing_jobs_sagemaker (List all SageMaker AI Processing Jobs)
• list_transform_jobs_sagemaker (List all SageMaker AI Transform Jobs)
• list_inference_recommender_jobs_sagemaker (List all SageMaker AI Inference Recommender Jobs)
• list_inference_recommender_job_steps_sagemaker (List all steps for a SageMaker AI Inference Recommender Job)
• describe_training_job_sagemaker (Describe a SageMaker AI Training Job)
• describe_processing_job_sagemaker (Describe a SageMaker AI Processing Job)
• describe_transform_job_sagemaker (Describe a SageMaker AI Transform Job)
• describe_inference_recommender_job_sagemaker (Describe a SageMaker AI Inference Recommender Job)
• stop_training_job_sagemaker (Stop a SageMaker AI Training Job)
• stop_processing_job_sagemaker (Stop a SageMaker AI Processing Job)
• stop_transform_job_sagemaker (Stop a SageMaker AI Transform Job)
• stop_inference_recommender_job_sagemaker (Stop a SageMaker AI Inference Recommender Job)

List of Tools for SageMaker AI Pipelines

• list_pipelines_sagemaker (List all SageMaker AI Pipelines)
• list_pipeline_executions_sagemaker (List all Pipeline Executions for a SageMaker AI Pipeline)
• list_pipeline_execution_steps_sagemaker (List all steps for a SageMaker AI Pipeline Execution)
• list_pipeline_parameters_for_execution_sagemaker (List all parameters for a SageMaker AI Pipeline Execution)
• describe_pipeline_sagemaker (Describe a SageMaker AI Pipeline)
• describe_pipeline_execution_sagemaker (Describe a SageMaker AI Pipeline Execution)
• describe_pipeline_definition_for_execution_sagemaker (Describe a SageMaker AI Pipeline Definition for Execution)
• start_pipeline_execution_sagemaker (Start a SageMaker AI Pipeline Execution)
• stop_pipeline_execution_sagemaker (Stop a SageMaker AI Pipeline Execution)
• delete_pipeline_sagemaker (Delete a SageMaker AI Pipeline)

List of Tools for SageMaker AI User Profiles and Spaces

• list_user_profiles_sagemaker (List all SageMaker AI User Profiles)
• list_spaces_sagemaker (List all SageMaker AI Spaces)

List of Tools for SageMaker AI MLflow Managed Tracking Servers

• list_mlflow_tracking_servers_sagemaker (List all MLflow Tracking Servers)
• create_mlflow_tracking_server_sagemaker (Create a new MLflow Tracking Server)
• create_presigned_mlflow_tracking_server_url_sagemaker (Create a presigned URL for an MLflow Tracking Server)
• describe_mlflow_tracking_server_sagemaker (Describe an MLflow Tracking Server)
• start_mlflow_tracking_server_sagemaker (Start an MLflow Tracking Server)
• stop_mlflow_tracking_server_sagemaker (Stop an MLflow Tracking Server)
• delete_mlflow_tracking_server_sagemaker (Delete an MLflow Tracking Server)

List of Tools for SageMaker AI Domains

• list_domains_sagemaker (List all SageMaker AI Domains)
• create_presigned_domain_url_sagemaker (Create a presigned URL for a SageMaker Domain)
• describe_domain_sagemaker (Describe a SageMaker AI Domain)
• delete_domain_sagemaker (Delete a SageMaker AI Domain)

List of Tools for SageMaker AI Models

• list_models_sagemaker (List all SageMaker AI Models)
• describe_model_sagemaker (Describe a SageMaker AI Model)
• delete_model_sagemaker (Delete a SageMaker AI Model)

List of Tools for SageMaker AI Model Cards

• list_model_cards_sagemaker (List all SageMaker AI Model Cards)
• list_model_card_export_jobs_sagemaker (List all SageMaker AI Model Card Export Jobs)
• list_model_card_versions_sagemaker (List all versions of a SageMaker AI Model Card)
• describe_model_card_sagemaker (Describe a SageMaker AI Model Card)
• delete_model_card_sagemaker (Delete a SageMaker AI Model Card)

List of Tools for SageMaker AI Apps

• list_apps_sagemaker (List all SageMaker AI Apps)
• create_app_sagemaker (Create a SageMaker AI App)
• create_presigned_notebook_instance_url_sagemaker (Create a presigned URL for a SageMaker Notebook Instance)
• describe_app_sagemaker (Describe a SageMaker AI App)
• describe_app_image_config_sagemaker (Describe a SageMaker AI App Image Config)
• delete_app_sagemaker (Delete a SageMaker AI App)
• delete_app_image_config_sagemaker (Delete a SageMaker AI App Image Config)

Security Considerations

• Use AWS IAM roles with appropriate permissions
• Store credentials securely
• Use temporary credentials when possible

License

This project is licensed under the Apache License, Version 2.0. See the LICENSE file for details.