GPT-5-Codex likes to play with tools unnecessarily

What version of Codex is running?

codex-cli 0.36.0

Which model were you using?

GPT-5-Codex-High

What platform is your computer?

macoS

What steps can reproduce the bug?

1. Expose MCP tools to model
2. It will keep calling them unnecessarily. They lose focus on the job at hand.

What is the expected behavior?

1. It should focus on job on hand and not call tools unncessarily

What do you see instead?

<img width="2804" height="1056" alt="Image" src="https://github.com/user-attachments/assets/d76b2c69-6f9b-4257-9b54-22a4be5550c1" />

Additional information

_No response_