-
Notifications
You must be signed in to change notification settings - Fork 41
Description
Summary
Clerk already uses Speakeasy to generate this Python SDK from an OpenAPI spec. Speakeasy also supports generating a Terraform provider as an output target from the same spec — meaning much of the scaffolding for a Clerk Terraform provider is already in place.
This would let teams manage Clerk configuration (applications, organizations, roles, permissions, etc.) as infrastructure-as-code alongside the rest of their stack.
Context
- Clerk roadmap item (221 upvotes, currently in backlog): https://feedback.clerk.com/roadmap?id=82b49526-b724-4811-b999-bee569b94143
- Speakeasy Terraform support: https://www.speakeasy.com/docs/terraform
Since clerk-sdk-python is already generated via Speakeasy from an OpenAPI spec (see .speakeasy/gen.yaml), adding a Terraform provider target to the same pipeline would be a relatively low-lift way to deliver on the roadmap item — no need to hand-author provider resources or maintain a separate API client.
What we'd use it for
We're building an IaC engine for Clerk (roles, permissions, organization settings) and currently drive it through the Python SDK. A Terraform/Pulumi provider would let us (and others) manage Clerk declaratively with proper state tracking, drift detection, and plan/apply workflows — the same way we manage GCP, Datadog, and other services.
Ask
Would the Clerk team consider adding a Terraform provider output to the existing Speakeasy generation pipeline? Happy to help test or contribute if there's interest.