The company says a coding agent can use it to select, train, optimize, and deploy models on Amazon SageMaker AI within the same natural-language workflow as other toolkit skills, including EC2 and Amazon Bedrock.
AWS describes the toolkit as the single supported integration point, replacing its earlier AWS Labs skills and plugins. The AI/ML skill is also pre-installed in SageMaker Studio JupyterLab for use with Kiro or another coding agent. AWS lists inference optimization, model customization and fine-tuning, serverful training, MLflow tracking, JumpStart deployment and recommendations, and processing jobs for data preparation and ETL among the supported areas.
The announcement was posted October 13, 2026. AWS says the skill has no additional charge; users pay for the AWS resources their agents consume. The source does not provide benchmark results, supported-agent compatibility details beyond Kiro or a coding agent of choice, usage limits, or a migration timetable for the previous plugins and skills.