We could add a return_direct feature to the Agent module (similar to LangGraph).

[Cursx]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:)
• Please do not modify this template :) and fill in all the required fields.

1. Is this request related to a challenge you're experiencing? Tell me about your story.

I am using the Agent module to generate ECharts visualizations and large Markdown tables. While Dify handles small datasets well, I encounter significant challenges when dealing with large datasets (e.g., sensor trend graphs with 2500+ data points).

Currently, the tool's output must pass through the LLM before being returned. For large structured data, this leads to:

Unnecessary Token Consumption: The LLM consumes a huge amount of tokens to process raw data that doesn't need interpretation.

High Latency: Processing large JSON strings slows down the response significantly.

Context Window Issues: Large datasets risk exceeding the LLM's context limit.

2. Additional context or comments

No response

3. Can you help us with this feature?

• I am interested in contributing to this feature.