divedra-workflow-run

To resume a failed workflow session, use session progress and status checks to diagnose the failure, then apply session resume, rerun, or continue operations to restart execution from a specific step or artifact boundary.

You can run deterministic mock scenarios by selecting a workflow and executing it with mock-scenario controls, ensuring local-only usage to avoid backend calls while validating the workflow bundle behavior.

Yes, you can inspect runtime and session details using GraphQL detail queries, which allow you to query merged execution states and logs for deep diagnostics during supervised long-running runs.

To run workflows against a remote GraphQL endpoint, you need to start the local control plane using serve, then pass the remote endpoint configuration to execute the workflow with server-backed execution and GraphQL inspection.

You can validate and inspect a workflow bundle by using workflow list to discover available bundles, workflow validate to check bundle integrity, and workflow inspect to view the normalized structure before execution.

Yes, supervised execution supports auto-improve options for long-running workflow runs, allowing the system to automatically apply improvements while monitoring progress and enabling session recovery if failures occur.