MCP Tools DirectoryAbout this integration
Amazon SageMaker AI MCP provides sagemaker resource inspection, sagemaker training-job discovery. Use the default mode without --allow-write or --allow-sensitive-data-access. Requires authorized SageMaker resources; job creation, deletion and log access are excluded.
- Transport
- stdio
- Authentication
- unknown
- Initial setup
- credentials
- Runtime
- unattended
- Evidence
- documented
- Package
- awslabs.sagemaker-ai-mcp-server
- Last compatibility test
- Not independently tested
Connect your agent
awslabs.sagemaker-ai-mcp-serverThe reviewed launcher enters the MCP request loop without an interactive terminal session. Use the default mode without --allow-write or --allow-sensitive-data-access. Requires authorized SageMaker resources; job creation, deletion and log access are excluded. Declared launcher: awslabs.sagemaker-ai-mcp-server. See the separately cited authentication and connection facts for prerequisites. Publisher documentation and source reviewed; package and endpoints not executed or independently tested. Initial provisioning and MCP client tool permissions must be configured by the operator.
Capabilities: SageMaker resource inspection, SageMaker training-job discovery
Connected profiles
Additional details
awslabs.sagemaker-ai-mcp-server
Source ↗ · Checked 2026-09-171.1.0
Source ↗ · Checked 2026-09-17The reviewed launcher enters the MCP request loop without an interactive terminal session. Use the default mode without --allow-write or --allow-sensitive-data-access. Requires authorized SageMaker resources; job creation, deletion and log access are excluded.
Source ↗ · Checked 2026-09-17Requires preconfigured AWS machine credentials and IAM permissions for the reviewed operations. Interactive SSO/MFA acquisition is excluded. Python requirement: >=3.10. Additional resource and access prerequisites are stated in the Unattended configuration fact.
Source ↗ · Checked 2026-09-17Use the declared awslabs.sagemaker-ai-mcp-server console executable from the awslabs.sagemaker-ai-mcp-server Python package with stdio. Apply the flags and configuration described in the Unattended configuration fact. The MCP client must permit the selected tools without interactive approval.
Source ↗ · Checked 2026-09-17Publisher documentation reviewed; package and integration endpoint not executed or independently security-audited.
Source ↗ · Checked 2026-09-17