Summary

• from sdk we pass workflow_context in case of streaming (step.py) when we do self.active_executor.run . This workflow context consists of things like step id, step name, step index etc which is then put in RunOutputEvent used by FE.
• Now in case of a custom function step, which does agent.run() inside it. this agent run is completely unanimous and not in the flow of our execution. like the user decides to trigger it. and we cant directly inject that workflow_context into this run, so - in case of this step even if we yield events the FE wont know about which step it belongs to because it doesn't have access to that step_id etc info.

Now we inspect all events that are yielded during the workflow run, and set the workflow context variables on them before yielding upwards.

• Also updates parallel step event streaming logic to immediately yield events as they come so that FE can catch and display in real time.
  • Old approach - collect all, yield at end
  • New approach - stream immediately (used queues as we cant yield directly from async running tasks)

Type of change

• Bug fix
• New feature
• Breaking change
• Improvement
• Model update
• Other:

Checklist

• Code complies with style guidelines
• Ran format/validation scripts (./scripts/format.sh and ./scripts/validate.sh)
• Self-review completed
• Documentation updated (comments, docstrings)
• Examples and guides: Relevant cookbook examples have been included or updated (if applicable)
• Tested in clean environment
• Tests added/updated (if applicable)

Additional Notes

Add any important context (deployment instructions, screenshots, security considerations, etc.)