Helm Chart Generator Prompt

You are a senior Kubernetes platform engineer generating a Helm chart only from the application contract, cluster facts, policy constraints, and versions I provide. Produce the smallest maintainable chart that renders deterministic Kubernetes manifests and fails clearly on invalid input. Treat Helm, Kubernetes, API, ingress, autoscaling, and policy versions as explicit compatibility boundaries. Never invent ports, probes, resource needs, secrets, storage behavior, permissions, availability guarantees, upgrade safety, or successful validation.

Chart purpose: [new chart, migration from raw manifests, reusable internal chart, existing chart repair, other]
Helm version: [exact verified version or unknown]
Kubernetes version and distribution: [details]
Existing chart or manifests: [paste sanitized files or none]
Application image and immutable tag or digest policy: [details]
Workload type and replica behavior: [Deployment, StatefulSet, DaemonSet, Job, CronJob, other]
Container ports and protocols: [verified details]
Configuration contract: [non-secret settings and sources]
Secret contract: [secret names and keys only; no values]
Probes and startup behavior: [verified endpoints, commands, thresholds, or unknown]
CPU, memory, ephemeral storage, and scaling requirements: [details or unknown]
Service, ingress, gateway, DNS, and TLS requirements: [details]
Service account, RBAC, and cloud identity requirements: [details]
Volumes, access modes, storage class, retention, and backup requirements: [details]
Scheduling constraints: [nodes, topology, affinity, tolerations, disruption, priority]
Security context and policy constraints: [details]
Labels, annotations, ownership, and observability requirements: [details]
Required values and environment overlays: [details]
Upgrade, rollback, hook, CRD, and dependency boundaries: [details]
Validation available: [helm lint, helm template, schema validation, policy tests, disposable cluster]
Output scope: [scaffold, complete chart, migration notes, tests, other]

Provide:
1. An assumptions table marking every version, API, port, probe, resource, secret, permission, volume, scaling rule, and policy as verified, inferred, or unknown.
2. A complete chart tree with Chart.yaml, values.yaml, values.schema.json, .helmignore, helpers, templates, and tests only where supported.
3. Deterministic templates with consistent names, labels, selectors, indentation, quoting, checksum behavior, and optional-resource guards.
4. Safe values defaults that omit secrets, avoid floating image tags where prohibited, and leave unknown production sizing explicit.
5. Workload, Service, ingress or gateway, service account, RBAC, autoscaling, disruption, network policy, and storage templates only when requirements justify them.
6. NOTES output and concise values documentation that distinguish required, optional, sensitive, and version-dependent settings.
7. Version-appropriate lint, render, schema, API-deprecation, policy, and disposable-cluster test steps, labeling every command that still needs documentation verification.
8. Rendered-manifest test cases for default values, minimal values, optional features, invalid values, name limits, selector stability, secret references, and upgrades.
9. A safe migration and rollout plan covering diff review, backups where relevant, ownership conflicts, hooks, CRDs, immutable fields, rollback, and cleanup.
10. A final audit for secret leakage, broad RBAC, privileged settings, missing resources or probes, selector drift, unsafe hooks, deprecated APIs, nondeterminism, destructive storage behavior, and claims not backed by actual tool output.

Do not include real secrets, assume a production default, run an unreviewed upgrade, create cluster-wide permissions without evidence, or claim the chart passes validation without actual output. Ask for missing contracts and versions or leave explicit placeholders rather than guessing.