add support for custom anthropic models

[udit-329]

What it does

Adds support for custom anthropic models. This works similar to how user can configure custom openAI models for situations such as when the model is hosted on a custom url.

Addresses #16666

How to test

If a user has an Anthropic model that is accessed from a custom endpoint, they can add it in settings.json similar to how it can already be done for openAI models.

"ai-features.AnthropicCustom.customAnthropicModels": [
        {
            "model": "claude-sonnet-4",
            "url": "<custom-url>",
            "id": "<model-ID>",
            "apiKey": "<API-Key>"
        }
    ]

Follow-ups

Breaking changes

• This PR introduces breaking changes and requires careful review. If yes, the breaking changes section in the changelog has been updated.

Attribution

Review checklist

• As an author, I have thoroughly tested my changes and carefully followed the review guidelines
• User-facing text is internationalized using the nls service (for details, please see the Internationalization/Localization section in the Coding Guidelines)

Reminder for reviewers

• As a reviewer, I agree to behave in accordance with the review guidelines

[sdirix]

Thanks for the contribution ❤️

I tested it successfully by defining a custom model against the official API endpoint 👍

However there is an issue with the "apiKey"-less path, see my comments.

[sdirix]

We don't update the year in the Theia project and just leave the header alone once a file is created.

[sdirix]

Suggested change

	export const CUSTOM_ENDPOINTS_PREF = 'ai-features.AnthropicCustom.customAnthropicModels';
	export const CUSTOM_ENDPOINTS_PREF = 'ai-features.anthropicCustom.customAnthropicModels';

[sdirix]

Suggested change

	markdownDescription: nls.localize('theia/ai/anthropic/useResponseApi/mdDescription',
	markdownDescription: nls.localize('theia/ai/anthropic/customEndpoints/mdDescription',

[sdirix]

This change is not sufficient. Within the AnthropicModel we check for the api key and if it does not exist, the model will immediately throw an error before even trying to use the SDK, see here.
If we allow apiKey-less models to report ready, then there must be a successful path for them.

Please check what is required for this. In the OpenAI SDK we always need to hand over a key, so for keyless models we hand over a "stub key". I don't know whether this is also required for the Anthropic SDK or whether it actually supports not handing over a key if the endpoint does not need one.

[ndoschek]

Hi @udit-329, just a gentle ping, the review is done and there are a few points that would need your input before we can proceed. TIA!