Summary

This PR implements a production-ready Anthropic Claude LLM integration for enhanced RCA summarization. The implementation includes automatic retry logic, token tracking, cost estimation, and comprehensive error handling.

Changes

• Implemented AnthropicLLM class with full Anthropic API integration
• Added exponential backoff retry logic (up to 3 attempts)
• Implemented token usage tracking and cost estimation
• Created structured system prompt for RCA summarization with:
  • Executive summary
  • Root cause analysis with confidence levels
  • Impact assessment
  • Step-by-step remediation
  • Monitoring recommendations
• Added enhance_remediation() method for LLM-powered remediation suggestions
• Exported AnthropicLLM and DummyLLM from package init.py
• Updated README with Anthropic LLM usage examples
• Added logging throughout for observability

Type of Change

• New feature (non-breaking change adding functionality)

Testing

Designed to work with:

• Claude 3.5 Sonnet (default model)
• API key from ANTHROPIC_API_KEY environment variable
• Graceful fallback to DummyLLM on errors

Features

• Retry Logic: Exponential backoff (1s, 2s, 4s) for API failures
• Token Tracking: Logs input/output tokens and cumulative usage
• Cost Estimation: Calculates costs based on Claude 3.5 Sonnet pricing
• Error Handling: Falls back gracefully with detailed logging
• Usage Stats: get_usage_stats() method for monitoring

Checklist

• Code follows project style guidelines
• Self-review completed
• Documentation updated
• No new warnings introduced

Related Issues

Implements Feature #6 from feature roadmap