Back to Cookbook

Helm Forge

Generate Helm charts without wrestling Go templates

Describe your application or hand it existing K8s manifests and get a proper Helm chart — values.yaml, templates with sensible defaults, and none of the Go template spaghetti. Also reviews and fixes existing charts.

House RecipeWork2 min

INGREDIENTS

🐙GitHub

PROMPT

Create a skill called "Helm Forge". Help me create and manage Helm charts: Mode 1 — Generate: Given a description of my application or existing Kubernetes YAML manifests, generate a complete Helm chart with: - Chart.yaml with proper metadata - values.yaml with sensible defaults and documentation comments - Templates for all resources, properly parameterized - NOTES.txt with post-install instructions - .helmignore Mode 2 — Review: Given an existing Helm chart, review for: - Best practices (resource limits, probes, security contexts) - Template quality (unnecessary complexity, missing conditionals) - values.yaml organization and documentation - Dependency management Run `helm lint` and `helm template --debug` to validate output. Avoid deeply nested Go template conditionals — keep templates readable.

How It Works

Helm charts are powerful but writing them is painful. Go template syntax

is write-only code for most people. This skill generates charts from

plain English or existing manifests, and can review existing charts for

best practices.

What You Get

  • Complete Helm chart from app description or existing K8s YAML
  • Properly parameterized values.yaml with sensible defaults
  • Clean Go templates (no deeply nested conditionals)
  • Chart.yaml with proper metadata and dependencies
  • NOTES.txt with useful post-install instructions
  • Helm lint and template validation

Setup Steps

  1. Either describe your app or provide existing K8s manifests
  2. Review the generated chart structure
  3. Run `helm lint` and `helm template` to validate
  4. Customize values.yaml for your environments

Tips

  • Start with existing manifests if you have them — conversion is more accurate than generation from scratch
  • Ask for environment-specific values files (values-dev.yaml, values-prod.yaml)
  • Request a helmfile.yaml if you're managing multiple releases
  • For existing charts, ask your Claw to review for best practices (resource limits, security contexts, probes)
Tags:#kubernetes#helm#devops#automation