New agent skill: Amazon SageMaker optimized generative AI inference for your coding agent

This article announces the aws-ai-ml skill for Amazon SageMaker AI, available through the Agent Toolkit for AWS, which gives coding agents expertise in inference optimization and benchmarking.

• Benchmark existing SageMaker endpoints and get performance metrics including throughput, latency, and concurrency
• Find optimal instance types for models stored in S3, JumpStart, or Hugging Face Hub
• Compare multiple benchmark runs to measure performance improvements or regressions
• Generate executable SageMaker Python SDK v3 code for all optimization tasks
• Install via Agent Toolkit for AWS or use pre-configured JupyterLab space in SageMaker Studio
• Works with any MCP-compatible coding agent including Kiro, Claude Code, and Codex

The skill bridges the gap between user intent and infrastructure by letting engineers describe their optimization goals in natural language and receive data-driven deployment recommendations.