v0.1.7 - Performance Evaluation Layer

Added

Performance Evaluation Layer

• Complete post-backtest performance analysis framework with extended risk metrics:
  • Stability metrics: rolling Sharpe, consistency scores, regime analysis
  • Risk-adjusted returns: Calmar, Sortino, profit factor, tail ratio
  • Return attribution: directional, signal strength, win/loss decomposition
• PerformanceConfig for immutable analysis parameters (frozen dataclass)
• PerformanceResult with structured performance summary and diagnostics
• PerformanceRegistry for tracking analysis history with JSON catalog
• Markdown report generation with comprehensive metrics and visualizations
• 5-step performance analysis workflow
• Production-ready notebook (notebooks/05_performance_analysis.ipynb) with 31 cells
• Comprehensive test suite with 6 test modules (136 tests total)
• Performance evaluation design documentation

Infrastructure Improvements

• Performance report directory structure in project config
• Enhanced backtest persistence with performance metrics in metadata
• Updated backtest notebook to save comprehensive performance data

Changed

• Removed examples folder and all references (replaced by production notebooks)
• Simplified .gitignore to focus on actual runtime artifacts
• Removed unnecessary .gitkeep files from data directories
• Updated evaluation layer __init__.py with performance module exports
• Enhanced governance design documentation with performance layer details

Fixed

• Corrected documentation claims about speculative features
• Improved alignment between design docs and actual implementation