ring-creating-helm-charts

name

ring:creating-helm-charts

description

Creating Helm charts to Lerian conventions via ring:helm: standardized chart structure, full env-var coverage from .env.example, security defaults (runAsNonRoot, readOnlyRootFilesystem), ClusterIP-only services, and health probes; validates helm lint and template render. Use when creating, modifying, or reviewing a chart, or migrating docker-compose to Helm. Skip for app-code-only changes or docker-compose-only deployments.

Helm Chart Creation & Maintenance

When to use

• Creating a new Helm chart for any Lerian service
• Modifying an existing Helm chart (adding components, dependencies, templates)
• Reviewing a Helm chart PR for convention compliance
• Migrating a docker-compose setup to Helm

Skip when

• Modifying only application code (no chart changes)
• Working on non-Helm deployment (docker-compose only) → use backend engineer via ring:implementing-tasks

Sequence

Standalone/on-demand. Not part of the lean backend dev-cycle.

Related

Complementary: ring:implementing-tasks

Standards reference: dev-team/docs/standards/helm/ Executor agent: ring:helm

You orchestrate. ring:helm creates chart files.

Step 1: Validate Input

Required: service_name, chart_type (single|multi-component|umbrella), components (non-empty). Optional: dependencies (postgresql, mongodb, rabbitmq, valkey, keda), has_worker, namespace.

Step 2: Naming Convention

Default: {service_name}-helm  (e.g., reporter-helm, tracer-helm)
Exceptions (no -helm suffix):
  - plugin-access-manager
  - otel-collector-lerian

Step 3: Dispatch Agent

Task:
  subagent_type: "ring:helm"
  description: "Create Helm chart for {service_name}"
  prompt: |
    ## Helm Chart Creation

    service_name: {service_name}
    components: {components}
    dependencies: {dependencies}
    chart_type: {chart_type}
    namespace: {namespace}

    Standards: Load dev-team/docs/standards/helm/ files.

    ## Required Steps
    1. Read application .env.example and bootstrap/config.go
       — extract ALL env vars (missing vars = CrashLoopBackOff)
    2. Verify health check endpoint in application source
    3. Create chart structure:

    charts/{service_name}-helm/
    ├── Chart.yaml
    ├── values.yaml
    ├── 
        
        
        
        
           
         
        
      

          
          
    
          
          
            
               
           
         
              
          

    
           
         
         

          

    
            
       
         
       

Step 4: Validate Output

if env_vars_missing > 0:
  → FAIL: list missing vars, re-dispatch

if helm lint fails:
  → Re-dispatch with specific lint errors

if all checks PASS:
  → Proceed to worker setup or final validation

Worker Chart (if has_worker = true)

Additional dispatch for worker component:

• Separate Deployment without Service
• Different resource limits (CPU-focused, no port exposure)
• Same configmap/secrets references
• LivenessProbe via process check (not HTTP)

Validation Checklist

## Helm Chart Validation

| Check | Status | Evidence |
|-------|--------|----------|
| Env var coverage (100%) | ✅/❌ | X/Y vars mapped |
| helm lint PASS | ✅/❌ | command output |
| helm template renders | ✅/❌ | YAML valid |
| Security context set | ✅/❌ | deployment.yaml:{line} |
| Service type = ClusterIP | ✅/❌ | service.yaml:{line} |
| Health probes match endpoints | ✅/❌ | deployment.yaml:{line} |
| No real secrets in values | ✅/❌ | |