Replace the (empty) sentinel with descriptive error messages:

TL;DR

Replace the (empty) sentinel in run_agent.py with descriptive error messages to improve user experience when the model returns empty content.

Guidance

• Identify the conditions under which the model returns empty content, such as reasoning-only responses or fully empty responses, to provide accurate error messages.
• Modify run_agent.py to set final_response to a descriptive error message instead of (empty), depending on the type of empty content returned by the model.
• Add defensive catches in gateway/run.py for both the main response path and background agent path to handle empty content scenarios.
• Test the updated run_agent.py and gateway/run.py with different types of model responses to ensure the error messages are correctly displayed to users.

Example

if model_response == "":
    final_response = "The model returned no content. Please try again."
elif model_response == "reasoning_only":
    final_response = "The model produced internal reasoning but no visible response."

Notes

The provided diff is from a different fork and needs to be ported to the current main branch. The exact implementation may vary depending on the specific requirements and constraints of the project.

Recommendation

Apply workaround: Replace the (empty) sentinel with descriptive error messages and add defensive catches in gateway/run.py to improve user experience and handle empty content scenarios.