helm_skill

Overview

This skill provides production-grade Helm 4 chart development, release management, and debugging workflows with security-first defaults. It auto-detects app details from Dockerfiles and code, generates library-pattern charts, and integrates OCI and GitOps workflows for reliable delivery. Use it to create hardened charts, manage releases atomically, and recover failed deployments quickly.

How this skill works

The skill analyzes project artifacts (Dockerfile, code, existing charts) to infer ports, health endpoints, users, and resource hints, then scaffolds library and application charts with values schema validation. It supports Helm v4 features (server-side apply, kstatus watching, OCI-first), provides commands for linting, templating, packaging, pushing to registries, and automates validation and policy checks. For operations it exposes release lifecycle actions (install/upgrade/rollback), value resolution tracing, hook inspection, and debugging workflows for failed releases.

When to use it

• Creating a new Helm chart or converting an app to a Helm-managed deployment
• Building reusable library charts and thin application charts (library pattern)
• Publishing charts to OCI registries and preparing charts for GitOps (ArgoCD/Flux)
• Managing releases with atomic install/upgrade and rollback strategies
• Debugging failed releases, hook failures, or manifest rendering issues
• Setting up environment overlays and values schema validation for prod readiness

Best practices

• Generate library charts for reusable templates and keep application charts thin
• Enable values.schema.json to validate configuration across environments
• Use server-side apply and kstatus-aware health checks to align with Helm 4
• Keep secure defaults: runAsNonRoot, readOnlyRootFilesystem, and minimal capabilities
• Pin OCI digests for immutable installs and version charts with SemVer
• Run helm lint, helm template --debug, dry-run installs and policy scans (OPA/Trivy) before release

Example use cases

• Auto-generate a production-hardened chart from a Dockerfile with detected probes and ports
• Create an umbrella chart to deploy a microservice suite with conditional dependencies
• Push a packaged chart to GHCR and configure ArgoCD to deploy from the registry
• Trace values resolution and recover a failed upgrade caused by hook timeouts
• Prepare charts for CI: lint, template render, dry-run, schema and policy validation