Skip to content

[FEATURE]: preserve model thinking effort setting when switching between primary agents with same model selected. #36703

Description

@IdrisGit

Feature hasn't been suggested before.

  • I have verified this feature I'm about to request hasn't been suggested before.

Describe the enhancement you want to request

When using the same model for multiple primary agents (e.g. Plan and Build), the thinking effort is currently preserved per model, not per agent. Sometimes we want to use the same model with different thinking levels depending on the agent.

One of my primary workflows is to use GPT models with high or xhigh thinking effort for the Plan agent to research ideas and understand/analyze the codebase. Where as I use the same GPT model with low thinking effort for the Build agent.

This works fine when I am using two different models (e.g GPT-5.6 and DeepSeek V4 Pro), since each model keeps its own setting. However, when both agents use the same LLM, the thinking effort always stays on the last selected value.

for example, if I switch from Plan to Build, the thinking effort remains on high, even though I want the Build agent to use low.

It would be great if thinking effort could be remembered per agent rather than only per model.

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Fields

No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions