Terramate and Terraform: Managing Cross-Stack Variable Propagation

Orchestrating Multi-Stack Dependencies

Terramate functions as an orchestration layer that generates Terraform-compatible files from templates defined in terramate.tm.hcl. When managing a complex infrastructure hierarchy, a common requirement is the propagation of output values from a base networking stack to multiple downstream application stacks.

Integration Constraints

While Terramate simplifies the distribution of common variables, it does not replace Terraform's native state management. The uncertainty lies in the most efficient way to handle dynamic dependencies—where a value produced by one stack must be consumed by another—without creating rigid coupling or manual update cycles during the terramate generate process.

Given a multi-stack architecture in Terramate v0.x, what is the recommended pattern for linking stack outputs to inputs across different directories to avoid circular dependencies? Should this be handled via Terramate's template variables or via Terraform remote state data sources?