You build the helm-chart skill first, then refine it through templating, testing, and distribution patterns so you can ship production-ready charts for your agents.
Each chapter ends with a skill reflection: test, find gaps, and improve the skill.
You finish with a production Helm chart for your Kubernetes-deployed agent plus a refined Helm skill. The sub-module follows the 4-Layer method: foundational templating → advanced composition → AI-assisted authoring → spec-driven capstone.