In the consuming workspace, navigate to Settings > Run Triggers and select the source workspaces whose successful applies should trigger a run in the consuming workspace.
Each workspace can subscribe to up to 20 source workspaces; when any source workspace completes a successful apply, the consuming workspace queues a new run automatically.
Use the tfe_workspace_run_trigger resource in the Terraform provider for HCP Terraform to manage run triggers as code rather than through the UI.
In the consuming workspace's configuration, use the tfe_outputs data source to read remote state outputs from the source workspace rather than using terraform_remote_state directly.
Test the trigger by applying a change in the source workspace and confirming a new run is queued in the consuming workspace without manual intervention.
Use workspace tagging and naming conventions to document the dependency graph, since HCP Terraform does not provide a visual dependency map for run triggers.
Known gotchas
Run triggers only fire on successful applies; if the source workspace apply fails or is discarded, the consuming workspace does not queue a run, which can leave dependent infrastructure out of sync.
The user configuring a run trigger must have admin access to the consuming workspace and read-runs permission on the source workspace; insufficient permissions result in a silent failure to save the trigger.
Run triggers are one-directional and do not detect cycles; misconfiguring bidirectional triggers between two workspaces causes an infinite run loop that must be manually stopped.
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp