h

Instruction file imported from cloudbees-oss/terraform-aws-cloudbees-ci-eks-addon (.github/instructions/h.instructions.md). Copyright stays with the author.

When working with Terraform code in this repository:

Terraform Standards

• Always include main.tf, variables.tf, outputs.tf, versions.tf in modules
• Use snake_case for resources, kebab-case for modules
• All variables and outputs MUST have descriptions
• Always tag resources with Name, Environment, Terraform = "true"

Code Quality

• Use 2-space indentation
• Group related resources with blank lines
• Add comments for complex logic
• Follow established file naming patterns in the project

AWS/EKS Best Practices

• Use least privilege for IAM roles
• Deploy across multiple AZs for HA
• Configure proper EKS authentication in providers

Security

• Never hardcode secrets or credentials
• Use AWS Secrets Manager for sensitive data
• Implement proper RBAC for Kubernetes
• Apply security groups with minimal access

Provider Syntax

• Helm provider: use kubernetes { } block syntax (NOT kubernetes = { })
• kubectl provider: use gavinbunney/kubectl not hashicorp/kubectl
• Configure EKS authentication with exec blocks using api_version "client.authentication.k8s.io/v1beta1"
• Use proper provider aliases for multi-region setups
• Helm provider versions >= 2.5.1 require block syntax for kubernetes configuration
• When Terraform reports "Blocks of type kubernetes are not expected here", check Helm provider version

Error Handling

• Use proper depends_on for resource dependencies
• Handle circular dependencies with data sources appropriately
• Implement lifecycle rules when resources need special handling

AI-Specific Guidelines

• When generating Terraform code, always include variable descriptions
• Suggest appropriate resource names following project conventions
• Include relevant tags for all AWS resources
• Prefer using existing module patterns over creating new ones
• When asked for fixes, explain the reasoning behind changes
• Consider disaster recovery implications for multi-region deployments
• Validate resource relationships and suggest missing dependencies
• Recommend security best practices for new resources