Current Situation:
When agents are executed in "Deep Research" mode via Automation, the resulting output includes comprehensive thinking and reasoning steps. In many use cases, these "reasoning traces" cause the total output to exceed the character limits supported by Jira Automation, leading to truncated responses or execution failures.

Suggestion:
To streamline downstream processing, it would be beneficial to have a configuration toggle that allows users to enable or disable the inclusion of reasoning steps in the final output. Providing a "clean" output option would eliminate the need for manual parsing or additional automation steps to strip reasoning traces, ensuring the content remains within character constraints and is ready for immediate use.