- Create build_query_params utility function in oura_api_client/utils
- Handles date conversion from date objects to ISO strings
- Filters out None values automatically
- Refactor all endpoint modules to use the new utility
- Reduces code duplication across 16 endpoint files
- All 101 tests passing

This centralizes the common pattern:
- Converting date objects to ISO format strings
- Building params dict with start_date, end_date, next_token
- Filtering out None values

Closes #9

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

- Fix undefined variable in ring_configuration.py
- Remove unused List import from webhook.py
- Add newline at end of query_params.py

- Migrate heartrate.py and personal.py from dataclasses to Pydantic BaseModel
- Fix critical formatting issues in daily_sleep.py with proper indentation
- Make sleep.py self-contained by defining its own contributor models
- Remove cross-file imports that created tight coupling between models
- Update test imports to use correct model types for each endpoint
- Maintain backward compatibility with from_dict() methods
- All 101 tests passing, flake8 clean

This standardizes all 20 model files to use Pydantic BaseModel consistently,
improving code maintainability and reducing coupling between modules.

Closes #10

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

Added a robust error handling system with:
- Custom exception hierarchy for specific HTTP error codes
- Automatic retry logic with exponential backoff for transient failures
- Rate limit handling with Retry-After header support
- Configurable retry behavior via RetryConfig
- Comprehensive test coverage for all error scenarios

Also added CLAUDE.md for project-specific instructions to maintain
consistency and best practices.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

- Update test to use specific ConnectionError instead of generic RequestException
- Fix flake8 issues by removing unused import
- Refactor parse_openapi.py to reduce complexity

Fix test incompatibilities with new error handling:
- Update 6 tests to use proper mock patterns with response.ok/status_code
- Replace deprecated raise_for_status.side_effect with create_api_error flow
- All previously failing tests now pass

Add parallel test execution for faster development:
- Add pytest-xdist to requirements-dev.txt
- Enable parallel testing with 'pytest -n auto' (14 workers on M4 Pro)
- Reduce test execution time from 2+ minutes to under 1 second

Enhance CLAUDE.md with debugging best practices:
- Add systematic test failure debugging approaches
- Document pytest strategies for isolation and targeted testing
- Include lessons learned about CI timeouts vs actual failures
- Update project commands to use parallel testing by default

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

- Add threshold-based re-reading (85%/25%/20-message rule) to minimize token waste
- Emphasize that mid-session context loss should be rare
- Include explicit uncertainty handling option
- Complete session continuity system documentation
- Update .gitignore to keep context files local and private

Technical architecture and direction by Gustavo Stor