feat: improve how to enable thinking mode

[CamilleLetavernier]

What it does

• introduce a new preference to configure the thinking mode (per provider, model or agent)
• introduce a new chat request setting to configure the thinking mode (for the current session)

How to test

• open the chat view
• configure the thinking mode in the { } menu
  • note: depending on the model/provider used, thinking mode may behave differently, so it's a bit hard to say precisely what to expect
  • this is especially chaotic for Ollama providers, because 1) not all models support thinking mode and 2) not all models interpret this option in the same way. Some thinking models may ignore the option entirely and will always output thinking tokens
  • general expectation: if thinking mode is enabled, tokens will be present; otherwise they should be absent

Regarding thinking tokens (budget), this concept is specific to Anthropic, but other providers may support a similar concept. For example, some OpenAI Thinking Models support "Low", "Medium" or "High" thinking settings, while others will ignore this setting entirely. The behavior may also depend on which OpenAI API version is used (according to preference ai-features.openAiOfficial.useResponseApi).

This feature is easiest to test with Anthropic models, as they (seem to) handle thinking mode more consistently than other providers.

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