Giving AI coding agents domain expertise
Skills and Powers are knowledge packages that give AI coding agents (Claude Code, Kiro) deep domain expertise. These packages provide battle-tested guidance derived from real deployment experience — no more guessing or hallucinated APIs.
Comprehensive Claude Code skill for Amazon SageMaker AI. Covers the full lifecycle: inference endpoints, training, HyperPod, Model Monitor, AutoGluon, and SDK v3 patterns. Consolidates and replaces three earlier skills.
View on GitHubA Kiro Power providing battle-tested guidance for deploying and training ML models on Amazon SageMaker AI. Pure knowledge base — no MCP servers required.
View on GitHubClaude Code skill for SageMaker LLM training — SFT, LoRA, QLoRA, DPO, and CPT patterns. Now consolidated into the unified sagemaker-ai skill.
View on GitHubAmazon SageMaker HyperPod skill for Claude Code — expert guidance for provisioning and managing HyperPod clusters. Now consolidated into the unified sagemaker-ai skill.
View on GitHubClaude Code skill for AWS HealthOmics genomics workflows.
View on GitHubAWS Forward-Deployed Engineer skill for AI/ML — builder-first technical partner for POCs with cost optimization and Well-Architected best practices.
View on GitHubClone the repo and add it as a local skill (Claude Code) or custom power (Kiro) — takes under a minute.
Ask your agent about deploying a model, fine-tuning an LLM, or setting up HyperPod. The relevant knowledge loads automatically.
Get correct SDK v3 code, validated container URIs, and production-ready patterns — no more guessing or hallucinated APIs.