[Enhancement]: Allow selecting Model Spec by URL Query Param

[leejss]

What features would you like to see added?

When accessing a page with an assistant_id, for example, via http://example.com/c/new?assistant_id=123456, it seems that the model selection section is not displaying the name defined in modelSpecs but is instead being set to null or undefined.

localize('com_ui_none_selected') is being displayed. It appears that the spec and iconURL in the conversation object are also being set to undefined.

Please revise it so that when accessing with the assistant_id query parameter, the name defined in modelSpecs is displayed.

More details

The modelSpecs I defined are in the following format.

modelSpecs:
  enforce: false
  prioritize: false
  list:
    - name: "Dall-E 3"
      label: "Dall-E 3"
      description: "Image Generator"
      iconURL: "fake_image.png"
      preset:
        endpoint: "assistants"
        assistant_id: "1234567"
        model: "gpt-4o"
    - name: "ServiceDesk"
      label: "Service Desk"
      description: "I am service desk"
      iconURL: "fake_icon.png"
      preset:
        name: "ServiceDesk"
        endpoint: "assistants"
        assistant_id: "12345678"
        model: "gpt-4o"

Which components are impacted by your request?

UI

Pictures

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

Hi, thanks for your suggestion. The assistant_id would not match your model spec because the system wouldn't know if it's associated to a spec or not.

I renamed the title to more accurately target the issue, and this would be good to have.

reference: https://www.librechat.ai/docs/features/url_query

[leejss]

Thank you for your comment and I have one more question.

I have multiple Open AI assistants in Librechat. I want to directly select a specific assistant through a URL query parameter, but the issue I faced was that when I selected one using the assistant_id query parameter, the model name and icon didn’t display properly. Is this problem inevitable? Or if there’s a way to handle multiple assistants, please let me know.