Implement multi-provider support for chat functionality and enhance configuration management

[Reion19]

All Submissions:

• [ ✅ ] Have you followed the guidelines in our Contributing document?
• [ ✅ ] Have you checked to ensure there aren't other open Pull Requests for the same update/change?
• [ ✅ ] Have you added an explanation of what your changes do and why you'd like us to include them?

Summary of Changes

This pull request includes three main commits:

Dependency Updates
Updated the gem dependencies to their latest stable versions.

RubyAI Quality-of-Life Improvements
Added .chat, .models, and .configure class methods to the RubyAI class. This allows configuration and usage without directly referencing RubyAI::Configuration.

Multi-Provider Support + Refactoring

    Refactored the gem’s configuration structure.

    Added support for multiple LLM providers using separate configuration parameters.

    Introduced the Chat class to allow instantiating different model instances and querying them via #call.

    Wrote specs for most new and changed methods.

    Added untested support for Anthropic LLMs (I couldn't test due to lack of access/funding).

    Made it easier to extend support for future providers.

What’s New?

You can now configure the gem "the Rails way" using RubyAI.configure do ... end.

Added a simple .chat method for sending prompts to the configured LLM.

The original usage with RubyAI::Client.new(...).call still works.

Introduced a Chat class for flexible use of different models.

Preliminary support for Anthropic models.

Easy-to-extend architecture for new LLM providers.

Let me know if you’d like tests for the Anthropic integration once I get access.

[a1xsh]

Adapter should come last

[Reion19]

its just an copy+paste from client, so i couldn't know that it should come last

[a1xsh]

Faraday can manage Json automatically

[Reion19]

the same thing as in the previous message

[a1xsh]

Why you stay with 1024 tokens? Can't we increase this?

[Reion19]

my bad, will be fixed in future commits, because in one of the next PR i've changed configuration of providers

I simply forgot about this thing.

[a1xsh]

Faraday will add this header automatically

[Reion19]

this line will be removed in future commits