invoke-ai/InvokeAI · error · ValueError

The selected saved workflow did not produce a valid workflow

Error message

The selected saved workflow did not produce a valid workflow_return output.

What it means

Once the single prepared workflow_return execution is located, its result is fetched from child_session.results and must be an instance of WorkflowReturnOutput. This ValueError is raised when the result is missing or of an unexpected type, so the parent cannot receive a valid return value from the child workflow.

Source

Thrown at invokeai/app/services/session_processor/workflow_call_runtime.py:157

        workflow_return_node_ids = [
            node_id for node_id, node in child_session.graph.nodes.items() if node.get_type() == "workflow_return"
        ]
        if not workflow_return_node_ids:
            raise ValueError("The selected saved workflow must contain exactly one workflow_return node.")
        if len(workflow_return_node_ids) > 1:
            raise ValueError("The selected saved workflow must not contain more than one workflow_return node.")

        workflow_return_node_id = workflow_return_node_ids[0]
        prepared_return_node_ids = child_session.source_prepared_mapping.get(workflow_return_node_id, set())
        if len(prepared_return_node_ids) != 1:
            raise ValueError(
                "The selected saved workflow produced an unsupported number of workflow_return executions."
            )

        prepared_return_node_id = next(iter(prepared_return_node_ids))
        output = child_session.results.get(prepared_return_node_id)
        if not isinstance(output, WorkflowReturnOutput):
            raise ValueError("The selected saved workflow did not produce a valid workflow_return output.")

        return output

    def resume_waiting_workflow_call(self, queue_item: SessionQueueItem) -> None:
        invocation = self.get_waiting_workflow_call_invocation(queue_item)
        child_session = queue_item.session.waiting_workflow_call_child_session
        if child_session is None:
            raise ValueError("Execution state is waiting on a workflow call but has no attached child session.")
        output = self.get_child_workflow_return_output(child_session)
        queue_item.session.end_waiting_on_workflow_call(status="completed")
        queue_item.session.complete(invocation.id, output)
        self._session_runner._on_after_run_node(invocation, queue_item, output)

    def fail_waiting_workflow_call(self, queue_item: SessionQueueItem, error_message: str) -> None:
        invocation = self.get_waiting_workflow_call_invocation(queue_item)
        queue_item.session.end_waiting_on_workflow_call(status="failed", error_message=error_message)
        self._session_runner._on_node_error(
            invocation=invocation,

View on GitHub (pinned to 0b6a024f2f)

Solutions

Ensure the child workflow runs to completion and the workflow_return node actually executes
Re-run / re-enqueue the workflow so results are regenerated
Check child session logs for an earlier node failure that prevented the return node from running
Upgrade InvokeAI if results were saved by an older version with a different output model

Example fix

// before: assuming the child completed
output = get_child_workflow_return_output(child_session)
// after: verify child completed first
if child_session.is_complete:
    output = get_child_workflow_return_output(child_session)

Defensive patterns

Strategy: validation

Validate before calling

output = child_session.results.get(prepared_return_node_id)
if output is None:
    raise RuntimeError("Child workflow finished without a workflow_return result; check child run for failures/cancellation.")

Type guard

from invokeai.app.invocations.output import WorkflowReturnOutput

def has_valid_return_output(session, prepared_return_node_id: str) -> bool:
    return isinstance(session.results.get(prepared_return_node_id), WorkflowReturnOutput)

Try / catch

try:
    output = runtime.get_child_workflow_return_output(child_session)
except ValueError as e:
    logger.warning("Child workflow_return missing/invalid: %s", e)
    # inspect child_session logs, then re-run the child workflow

Prevention

Never cancel/fail a child workflow before its workflow_return node executes
Check child session status is complete before extracting return output
Match InvokeAI versions between save and resume times (output schema stability)
Monitor child session logs for upstream node failures

When it happens

Trigger: child_session.results.get(prepared_return_node_id) returns None (the return node never produced a result, e.g. the child was canceled or failed before the return node ran) or stores an object that is not WorkflowReturnOutput when the parent resumes.

Common situations: Canceling or failing the child workflow before the return node executes, resuming a partially executed child session, or version mismatches where stored results were serialized under a different output schema.

Related errors

AI-assisted analysis of invoke-ai/InvokeAI@0b6a024f2f (2026-08-29). Data as JSON: /api/errors/c95222e621896c8d. Report an issue: GitHub.