Opencode: More detail in thinking messages

[ahostetler300]

I get much better detail in my thinking messages when using Opencode directly, compared to the thinking output I get in Maestro. Is this something you have looked at before?

This is a pretty basic example, but when actively developing or troubleshooting, these detailed thinking messages can be insightful. I do sometimes get more than this, but not consistently, and not nearly at the level I see in OpenCode.

Maestro

OpenCode

[chr1syy]

Let me look into how Thinking stream is generated. I know that claude-code does give out more informations during thinking-streams. I'll look into it.