/thinking incorrectly reports that GPT-5.4 does not support thinking/reasoning

[screenfluent]

I'm using GPT-5.4, but when I run /thinking and choose an option like high, kit shows:

Info
Current model does not support thinking/reasoning.

Expected behavior:

• /thinking should work for GPT-5.4
• If support depends on the exact model or provider, the detection should correctly identify that

Actual behavior:

• I select /thinking
• I choose a level such as high
• kit responds with:
  Current model does not support thinking/reasoning.

Environment:

• macOS 15.6.1

• Apple Silicon

• kit installed with:

  go install github.com/mark3labs/kit/cmd/kit@latest

[mark-iii-labs-huly]

Connected to Huly®: KIT-11