Isolated Convex Dev Deployments per Git Worktree / Branch

I frequently work on the same application using multiple Git worktrees in parallel. Currently, those worktrees may all be connected to the same Convex development deployment/database. This makes parallel development difficult because changes from one worktree can affect the shared Convex dev environment used by the others. For example, different worktrees may be modifying functions, schemas, indexes, or other backend code independently, but deploying from one worktree changes the environment that every other worktree is using. It would be very useful to have isolated Convex development deployments/databases that can be associated with individual worktrees or branches. Ideally, the workflow would look something like: Create a Git worktree/branch for a feature. Give that worktree its own isolated Convex dev deployment and database. Run convex dev and make schema/function/backend changes without affecting other worktrees. Have multiple worktrees running simultaneously, each against its own independent Convex environment. When a branch is merged, only the code/schema changes that are actually merged eventually get deployed to the shared/production Convex deployment. The temporary development deployment could then be deleted when the worktree/branch is no longer needed. Essentially, I’m looking for a workflow similar to preview environments for backend development, where each branch/worktree can have its own Convex deployment and data rather than multiple local development environments competing over one shared dev deployment. If Convex already supports this workflow, it would be great to have clearer tooling/documentation around setting it up. Otherwise, isolated or ephemeral dev deployments per branch/worktree would make parallel development—especially with multiple agents or developers working on the same project—much easier.