Assistant configuration is not being sent to AI provider when run as agent

Problem/Motivation

In the latest 1.2.x, as far as I can tell, the configured max tokens and temperature settings for an assistant are never sent to the AI provider when running as an agent. Unless I'm missing something (entirely possible).

Steps to reproduce (required for bugs, but not feature requests)

Please provide information like AI modules enabled, which AI provider, browser, etc.

Set max output tokens and temperature settings on an assistant that is configured to run as an agent.
Run the agent and verify that the max output and temperature settings have no effect.

Proposed resolution

Ensure that assistant configuration settings max tokens and temperature are communicated to the AI provider.

Remaining tasks

For assistants run as agents, ensure specified provider settings on the assistant are communicated to the AI provider.

Optional: Other details as applicable (e.g., User interface changes, API changes, Data model changes)