Add support for Anthropic Claude LLM

[M-Tahar]

Proposal to add Claude/Anthropic as a supported LLM provider in the bee-agent-framework.

Motivation:

• Claude offers strong performance and capabilities that would benefit Bee Stack users
• Adds more provider options for users alongside existing OpenAI, watsonx, and ollama options
• Complements existing providers with Claude's specific strengths (longer context windows, strong coding abilities, etc.)
• Provides an alternative for users who prefer Claude's approach to AI safety and alignment

Implementation Plan:

1. Add Claude adapter in src/adapters/claude

   • Implement API integration using Anthropic's official SDK
   • Support both Claude 2 and Claude 3 models
   • Handle Claude-specific features like system prompts and tool use

2. Update configuration:

   • Add Claude-specific environment variables (ANTHROPIC_API_KEY)
   • Add Claude as a supported backend option in bee-stack setup
   • Add model configuration options (temperature, max tokens, etc.)

3. Documentation:

   • Add setup instructions for Claude API
   • Document Claude-specific features and configurations
   • Update example configurations

4. Testing:

   • Add unit tests for Claude adapter
   • Add integration tests
   • Add example usage in documentation

Technical Details:

• Will follow existing adapter patterns seen in other providers
• Will implement proper error handling for Claude-specific API responses
• Will ensure compatibility with bee-stack's existing monitoring and observability features

Would love to contribute this feature if the maintainers are interested! Let me know if you have any specific requirements or preferences for the implementation.

[Tomas2D]

Hello, @M-Tahar, and thank you for your interest in Bee.

To start, please first go through our Contributing Guidlines followed by Adding a new LLM provider (adapter) instructions(https://i-am-bee.github.io/bee-agent-framework/#/llms?id=adding-a-new-provider-adapter). Feel free to open a draft PR where we can discuss.

Relevant PRs (adding a new LLM provider):

• Add GCP VertexAI adapter #164
• feat(adapters): add aws bedrock #168