Feature: Auto-expand sub-agent thinking in conversation, then auto-collapse when done

Resolved 💬 2 comments Opened Apr 1, 2026 by uncle-zhongyusun Closed May 8, 2026

Feature Request

Desired Behavior

  1. While thinking: Sub-agent thinking content should automatically expand and be visible inline in the conversation, so users can see the reasoning process without any clicks.
  2. After thinking completes: The thinking block should automatically collapse, keeping the conversation clean and readable.

Current Behavior

  • Sub-agent thinking is collapsed by default
  • Users must manually click to expand and view thinking content

Use Case

When running complex multi-step tasks with sub-agents, it's useful to observe the thinking process in real-time without having to click into each thinking block individually.

Proposed UX

  • Thinking block auto-expands as soon as a sub-agent starts thinking
  • Content streams in live
  • Once thinking is complete, the block auto-collapses back to the summary view

View original on GitHub ↗

This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗